Prompts

When to use

Establishes the project boundary, intended outcome, operating mode, authorization, protected surfaces, constraints, and unresolved assumptions.

Inputs and prerequisites

• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• .prompt_suite/control/project_context_and_run_configuration.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• core
• control
• context

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Defines mandatory authorization, evidence, trust, change-control, rollback, verification, and honesty rules for every run.

Inputs and prerequisites

• MD-00
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• .prompt_suite/control/universal_safety_authorization_and_evidence_contract.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• core
• control
• contract

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Selects the smallest coherent capability graph, resolves prerequisites and conflicts, and schedules safe investigation and execution waves.

Inputs and prerequisites

• MD-00
• MD-01
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• .prompt_suite/control/capability_router_and_execution_graph.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• core
• control
• orchestrator

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Defines typed evidence, findings, action plans, approvals, dry runs, execution logs, verification results, and residual records.

Inputs and prerequisites

• MD-00
• MD-01
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• .prompt_suite/control/artifact_handoff_and_verification_contract.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• core
• control
• contract

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Defines the suite tag grammar, placeholder syntax, instruction hierarchy, untrusted-data isolation, and prompt-injection resistance rules.

Inputs and prerequisites

• MD-00
• MD-01
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• .prompt_suite/control/prompt_structure_and_input_trust_contract.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• core
• control
• contract

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Frames the real problem, opportunity, actors, constraints, hypotheses, options, and decision criteria without mutating project state.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• reports/problem_opportunity_and_strategy_framing/problem_opportunity_and_strategy_framing_investigation.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• strategy
• investigative
• analysis
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Synthesizes user needs, market conditions, alternatives, competitors, and evidence quality into decision-ready findings.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• reports/market_user_and_competitive_research/market_user_and_competitive_research_investigation.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• strategy
• investigative
• research
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Converts an intended outcome into coherent functional, non-functional, operational, security, and acceptance requirements.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• reports/requirements_scope_and_success_criteria/requirements_scope_and_success_criteria_investigation.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• strategy
• investigative
• specification
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Evaluates feasibility, validates risky assumptions, defines prototype evidence, and creates a dependency-aware roadmap.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• reports/feasibility_prototype_risk_and_roadmap/feasibility_prototype_risk_and_roadmap_investigation.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• strategy
• investigative
• planning
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Evaluates architecture, boundaries, data flows, interfaces, and technology choices against explicit quality attributes.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• reports/architecture_and_technology_decision_analysis/architecture_and_technology_decision_analysis_investigation.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• engineering
• investigative
• decision_record
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Builds a factual inventory of code, configuration, data, infrastructure, documentation, ownership, workflows, and current health.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• reports/project_discovery_inventory_and_baseline/project_discovery_inventory_and_baseline_investigation.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• engineering
• investigative
• discovery
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Produces evidence-backed pitches, proposals, cases, and decision documents tailored to a specific audience, ask, risk, and evaluation process.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/pitch_proposal_and_business_case_development/pitch_proposal_and_business_case_development_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• APPLY_APPROVED
• VERIFY_ONLY

Tags and routing hints

• business
• operational
• generation
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Designs the audience journey, argument, evidence, visual grammar, slide sequence, interaction, and delivery plan before slide production.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• reports/presentation_narrative_story_architecture_and_deck_blueprint/presentation_narrative_story_architecture_and_deck_blueprint_brief.md
• typed_runtime_artifacts
• plan_review_package
• execution_consent_request

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• visual_communication
• investigative
• creative_design
• hybrid

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Builds a responsive, accessible, branded browser-native slide deck from an approved narrative blueprint, with optional export packaging.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• MD-101
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/html_css_and_javascript_slides_and_presentation_production/html_css_and_javascript_slides_and_presentation_production_package.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• APPLY_APPROVED
• VERIFY_ONLY

Tags and routing hints

• visual_communication
• executive
• visual_generation
• hybrid

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Builds an analytical or operational dashboard from verified data definitions, user decisions, interaction requirements, and performance constraints.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/dashboard_and_analytical_interface_production/dashboard_and_analytical_interface_production_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• APPLY_APPROVED
• VERIFY_ONLY

Tags and routing hints

• data_visualization
• operational
• interactive_generation
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Produces a research-backed infographic or visual data story with accurate scales, narrative hierarchy, annotations, sources, and accessible alternatives.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/infographic_and_data_storytelling_production/infographic_and_data_storytelling_production_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• APPLY_APPROVED
• VERIFY_ONLY

Tags and routing hints

• data_visualization
• operational
• visual_generation
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Creates implementation-ready blueprints, architecture diagrams, process maps, system views, and technical visual specifications using semantic HTML and SVG where appropriate.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/blueprint_diagram_architecture_and_process_visualization_production/blueprint_diagram_architecture_and_process_visualization_production_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• APPLY_APPROVED
• VERIFY_ONLY

Tags and routing hints

• visual_communication
• operational
• visual_generation
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Creates editable collaborative diagrams for workshops, architecture discussions, journeys, mapping, and ideation with clear grouping and facilitation logic.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/excalidraw_and_collaborative_whiteboard_diagram_production/excalidraw_and_collaborative_whiteboard_diagram_production_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• APPLY_APPROVED
• VERIFY_ONLY

Tags and routing hints

• visual_communication
• operational
• visual_generation
• hybrid

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Builds a self-contained responsive web artifact, explainer, microsite, prototype, calculator, or interactive narrative from an approved content and interaction specification.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/interactive_web_artifact_microsite_and_prototype_production/interactive_web_artifact_microsite_and_prototype_production_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• APPLY_APPROVED
• VERIFY_ONLY

Tags and routing hints

• web_design
• operational
• interactive_generation
• hybrid

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Transforms a functional frontend into a coherent, distinctive, accessible, responsive experience without changing product intent or hiding usability problems.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/frontend_visual_design_interaction_and_experience_polish/frontend_visual_design_interaction_and_experience_polish_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• APPLY_APPROVED
• VERIFY_ONLY

Tags and routing hints

• web_design
• operational
• transformation
• hybrid

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Reviews slides, dashboards, diagrams, infographics, web artifacts, and branded assets for communication quality, accessibility, responsiveness, consistency, and production defects.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• reports/visual_quality_accessibility_brand_and_responsive_review/visual_quality_accessibility_brand_and_responsive_review_brief.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• visual_quality
• investigative
• review
• hybrid

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Models assets, actors, entry points, trust boundaries, abuse cases, threats, controls, and residual exposure without performing attacks.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• reports/threat_modeling_and_attack_surface_analysis/threat_modeling_and_attack_surface_analysis_investigation.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• security
• investigative
• threat_model
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Designs an event from objective through experience, agenda, run-of-show, roles, logistics, content, risk, communications, accessibility, and measurement.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/event_strategy_agenda_run_of_show_and_facilitation_design/event_strategy_agenda_run_of_show_and_facilitation_design_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• APPLY_APPROVED
• VERIFY_ONLY

Tags and routing hints

• events
• operational
• planning
• hybrid

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Creates a dependency-aware operating plan spanning outcomes, workstreams, milestones, ownership, capacity, risk, governance, communication, and adaptation.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• reports/program_project_and_operations_planning/program_project_and_operations_planning_brief.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• planning
• investigative
• planning
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Designs focused meetings and workshops with clear decisions, preparation, activities, facilitation, artifacts, inclusion, and follow-through.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/workshop_meeting_and_collaborative_session_design/workshop_meeting_and_collaborative_session_design_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• APPLY_APPROVED
• VERIFY_ONLY

Tags and routing hints

• facilitation
• operational
• planning
• hybrid

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Produces clear, evidence-backed professional documents with audience-fit structure, decision logic, responsibilities, controls, and usable procedures.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/document_memo_report_sop_and_policy_production/document_memo_report_sop_and_policy_production_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• APPLY_APPROVED
• VERIFY_ONLY

Tags and routing hints

• professional_writing
• operational
• generation
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Develops a research-backed go-to-market strategy connecting audience, category, positioning, offer, channels, journey, sales, launch, economics, and learning.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• reports/marketing_audience_and_go_to_market_strategy/marketing_audience_and_go_to_market_strategy_brief.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• marketing
• investigative
• strategy
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Produces accurate, audience-specific sales and customer materials across discovery, evaluation, onboarding, adoption, support, renewal, and expansion.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/sales_enablement_customer_communication_and_lifecycle_content/sales_enablement_customer_communication_and_lifecycle_content_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• APPLY_APPROVED
• VERIFY_ONLY

Tags and routing hints

• sales_enablement
• operational
• generation
• hybrid

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Designs a transparent spreadsheet or financial model with inputs, formulas, scenarios, controls, auditability, outputs, and decision-focused visualizations.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/spreadsheet_financial_forecast_and_scenario_model_specification/spreadsheet_financial_forecast_and_scenario_model_specification_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• APPLY_APPROVED
• VERIFY_ONLY

Tags and routing hints

• analytics
• operational
• generation
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Designs an evidence-informed learning system covering learners, outcomes, prerequisites, sequence, practice, feedback, assessment, accessibility, and delivery constraints.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• reports/learning_experience_curriculum_course_lesson_and_assessment_research_and_design_brief/learning_experience_curriculum_course_lesson_and_assessment_research_and_design_brief_brief.md
• typed_runtime_artifacts
• plan_review_package
• execution_consent_request

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• learning_design
• investigative
• paired_investigation
• hybrid

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Produces aligned lessons, tutorials, activities, assessments, facilitator materials, learner resources, and delivery assets from the frozen learning design.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• MD-117
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/learning_experience_curriculum_course_lesson_and_assessment_material_production_and_validation/learning_experience_curriculum_course_lesson_and_assessment_material_production_and_validation_package.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• APPLY_APPROVED
• VERIFY_ONLY

Tags and routing hints

• learning_design
• executive
• paired_execution
• hybrid

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Designs valid, ethical, unbiased, usable instruments for interviews, surveys, observation, diary studies, assessments, and structured data collection.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• reports/survey_interview_questionnaire_and_research_instrument_design/survey_interview_questionnaire_and_research_instrument_design_brief.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• research
• investigative
• research_design
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Runs bounded, non-destructive, explicitly authorized attack simulations to validate defenses and produce reproducible evidence.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/authorized_attack_simulation_and_adversarial_validation/authorized_attack_simulation_and_adversarial_validation_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_APPROVED
• VERIFY_ONLY

Tags and routing hints

• security
• operational
• controlled_validation
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Produces evidence-backed grant, funding, sponsorship, fellowship, award, and institutional proposal packages aligned to eligibility, evaluation criteria, budget, impact, and compliance.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/grant_funding_sponsorship_and_award_proposal_production/grant_funding_sponsorship_and_award_proposal_production_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• APPLY_APPROVED
• VERIFY_ONLY

Tags and routing hints

• business
• operational
• publication_generation
• hybrid

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Creates truthful, role-specific resumes, CVs, bios, portfolios, cover letters, applications, and professional profiles from verified experience and evidence.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/professional_profile_resume_portfolio_bio_and_application_package/professional_profile_resume_portfolio_bio_and_application_package_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• APPLY_APPROVED
• VERIFY_ONLY

Tags and routing hints

• professional_communication
• operational
• generation
• hybrid

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Produces clear, audience-fit professional and public communications with the right action, tone, context, evidence, channel, and risk controls.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/email_letter_message_speech_announcement_and_general_communication_production/email_letter_message_speech_announcement_and_general_communication_production_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• APPLY_APPROVED
• VERIFY_ONLY

Tags and routing hints

• professional_communication
• operational
• generation
• hybrid

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Produces platform-fit short-form posts, threads, captions, community prompts, carousels, clips, and response packages from verified source material and an approved content strategy.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/social_media_short_form_and_community_content_production/social_media_short_form_and_community_content_production_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• APPLY_APPROVED
• VERIFY_ONLY

Tags and routing hints

• content_operations
• operational
• publication_generation
• hybrid

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Implement an approved bounded feature or integration across code, tests, interfaces, data, documentation, telemetry, and acceptance evidence without expanding scope silently.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/feature_change_and_integration_implementation/feature_change_and_integration_implementation_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• VERIFY_ONLY
• APPLY_APPROVED

Tags and routing hints

• engineering
• operational
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Design decision-relevant experiments with explicit hypotheses, randomization, metrics, power assumptions, guardrails, stopping rules, instrumentation, and interpretation limits.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/experiment_a_b_test_and_causal_measurement_design/experiment_a_b_test_and_causal_measurement_design_result.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• analytics
• investigative
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Transform content across languages and cultures while preserving meaning, terminology, claims, voice, accessibility, formatting, and legal or brand constraints.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/translation_localization_and_transcreation/translation_localization_and_transcreation_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• VERIFY_ONLY

Tags and routing hints

• localization
• operational
• hybrid

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Convert an approved script and asset inventory into an editing, motion, audio, caption, rights, export, channel, and quality-control blueprint without generating imagery.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/media_production_editing_and_distribution_blueprint/media_production_editing_and_distribution_blueprint_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• VERIFY_ONLY

Tags and routing hints

• media_production
• operational
• hybrid

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Analyze content performance, audience behavior, confounders, distribution effects, and editorial quality to propose testable improvements rather than vanity-metric conclusions.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/content_performance_and_editorial_optimization/content_performance_and_editorial_optimization_result.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• content_operations
• investigative
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Review a scholarly manuscript for contribution, methods, statistics, evidence, reproducibility, ethics, citations, and venue fit, then prepare constructive reviewer comments or an evidence-backed rebuttal.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/academic_peer_review_methodological_critique_and_rebuttal/academic_peer_review_methodological_critique_and_rebuttal_result.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• academic
• investigative
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Reproduces a specific suspected vulnerability in an isolated or explicitly approved environment and validates that a patch blocks it.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/safe_exploit_reproduction_and_patch_validation/safe_exploit_reproduction_and_patch_validation_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_APPROVED
• VERIFY_ONLY

Tags and routing hints

• security
• operational
• controlled_validation
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Improve findability through search intent, information architecture, internal linking, structured metadata, technical discoverability, answer quality, and anti-spam safeguards.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/search_discoverability_seo_aeo_and_information_architecture/search_discoverability_seo_aeo_and_information_architecture_result.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• marketing
• investigative
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Define requirements, evaluate vendors, compare total cost and risk, conduct security and privacy diligence, score proofs of concept, and produce an auditable recommendation and exit plan.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/procurement_vendor_evaluation_rfp_and_due_diligence/procurement_vendor_evaluation_rfp_and_due_diligence_result.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• procurement
• investigative
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Translate organizational demand into workforce capacity, role mix, skills, location, cost, succession, and scenario plans while separating evidence from sensitive employment decisions.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/workforce_strategy_and_capacity_planning/workforce_strategy_and_capacity_planning_result.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• people
• investigative
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Create fair role scorecards, sourcing criteria, structured interviews, work samples, evaluation rubrics, decision records, and onboarding plans with human authority and bias controls.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/hiring_interview_selection_and_onboarding_design/hiring_interview_selection_and_onboarding_design_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• VERIFY_ONLY
• APPLY_APPROVED

Tags and routing hints

• people
• operational
• hybrid

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Design transparent competency, expectation, feedback, growth, calibration, and learning systems that are role-relevant, inclusive, and reviewable by accountable humans.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/performance_competency_career_and_learning_frameworks/performance_competency_career_and_learning_frameworks_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• VERIFY_ONLY
• APPLY_APPROVED

Tags and routing hints

• people
• operational
• hybrid

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Design support taxonomy, intake, response playbooks, escalation, service recovery, knowledge feedback, quality review, staffing signals, and safe automation boundaries.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/customer_support_and_service_operations/customer_support_and_service_operations_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• VERIFY_ONLY

Tags and routing hints

• customer_operations
• operational
• hybrid

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Synthesize adoption, retention, health, feedback, outcomes, risks, and lifecycle signals into prioritized customer-success interventions and product learning.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/customer_success_retention_and_voice_of_customer/customer_success_retention_and_voice_of_customer_result.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• customer_operations
• investigative
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Prepare principled negotiations through interests, constraints, alternatives, objective criteria, issue packages, concession boundaries, questions, scenarios, and decision checkpoints without manipulation.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/negotiation_and_decision_preparation/negotiation_and_decision_preparation_result.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• strategy
• investigative
• hybrid

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Create a maintainable personal operating system for goals, projects, tasks, notes, decisions, learning, reviews, and information flow without unnecessary complexity.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/personal_knowledge_productivity_and_work_system/personal_knowledge_productivity_and_work_system_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• VERIFY_ONLY

Tags and routing hints

• personal_productivity
• operational
• hybrid

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Define a lawful, ethical, purpose-limited OSINT question, collection plan, source classes, privacy boundaries, preservation rules, stop conditions, and reporting standard.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/osint_research_planning_authorization_and_collection_governance/osint_research_planning_authorization_and_collection_governance_result.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• osint
• investigative
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Performs authorized adversarial evaluation of models and AI systems across misuse, leakage, injection, tool abuse, robustness, and control failure.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/model_and_ai_system_red_team_evaluation/model_and_ai_system_red_team_evaluation_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_APPROVED
• VERIFY_ONLY

Tags and routing hints

• model_security
• operational
• controlled_validation
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Collect authorized public information with provenance, timestamps, source validation, contradiction tracking, chain of custody, privacy minimization, and reproducible preservation.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/osint_collection_source_validation_and_preservation/osint_collection_source_validation_and_preservation_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• VERIFY_ONLY
• APPLY_APPROVED

Tags and routing hints

• osint
• operational
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Resolve entities, events, locations, relationships, and uncertainty into traceable timelines and networks while preventing identity conflation and unsupported attribution.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/osint_entity_timeline_geospatial_and_network_analysis/osint_entity_timeline_geospatial_and_network_analysis_result.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• osint
• investigative
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Produce estimative intelligence with key judgments, confidence, indicators, alternative hypotheses, assumptions, source limitations, warning thresholds, and decision implications.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/intelligence_estimate_warning_and_executive_brief/intelligence_estimate_warning_and_executive_brief_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• VERIFY_ONLY
• APPLY_APPROVED

Tags and routing hints

• intelligence
• operational
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Research legal issues using controlling and persuasive authority, jurisdiction and date limits, counterarguments, factual dependencies, uncertainty, and non-lawyer review boundaries.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/legal_research_and_issue_analysis/legal_research_and_issue_analysis_result.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• legal
• investigative
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Review contracts or policies for obligations, rights, ambiguity, inconsistency, risk allocation, privacy, security, operations, remedies, and negotiation points without pretending to provide final legal approval.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/contract_policy_and_clause_review/contract_policy_and_clause_review_result.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• legal
• investigative
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Analyze actuals, budgets, forecasts, drivers, unit economics, scenarios, variances, and management actions with reconciled definitions and explicit assumptions.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/finance_fp_a_budget_and_management_analysis/finance_fp_a_budget_and_management_analysis_result.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• finance
• investigative
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Design or review close procedures, reconciliations, journal support, evidence, segregation, exceptions, and control sign-offs without posting entries beyond authority.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/accounting_close_reconciliation_and_control_review/accounting_close_reconciliation_and_control_review_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• VERIFY_ONLY
• APPLY_APPROVED

Tags and routing hints

• finance
• operational
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Model cash timing, liquidity, covenant headroom, funding needs, concentration, downside scenarios, and decision triggers using reconciled assumptions.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/treasury_cash_flow_liquidity_and_runway_planning/treasury_cash_flow_liquidity_and_runway_planning_result.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• finance
• investigative
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Plan and perform risk-based control testing with criteria, samples, evidence, exceptions, root causes, management responses, and independent assurance boundaries.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/internal_audit_control_testing_and_assurance/internal_audit_control_testing_and_assurance_result.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• audit
• investigative
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Map an operating process, identify demand, delays, defects, controls, handoffs, capacity, and automation opportunities, then design a measurable improved state.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/operations_process_design_and_continuous_improvement/operations_process_design_and_continuous_improvement_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• VERIFY_ONLY

Tags and routing hints

• operations
• operational
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Coordinates evidence-preserving response to an active or suspected incident through triage, containment, eradication, recovery, and lessons capture.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/incident_triage_containment_eradication_and_recovery/incident_triage_containment_eradication_and_recovery_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_APPROVED
• VERIFY_ONLY

Tags and routing hints

• operations
• operational
• incident_response
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Analyze suppliers, inventory, lead times, service, logistics, capacity, concentration, disruption scenarios, controls, and resilience options.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/supply_chain_inventory_logistics_and_resilience_planning/supply_chain_inventory_logistics_and_resilience_planning_result.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• supply_chain
• investigative
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Connect user evidence, market context, business strategy, product outcomes, options, sequencing, lifecycle decisions, and roadmap tradeoffs.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/product_discovery_strategy_roadmap_and_lifecycle/product_discovery_strategy_roadmap_and_lifecycle_result.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• product
• investigative
• hybrid

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Produce implementation-ready product requirements, flows, edge cases, non-functional constraints, analytics, rollout, acceptance tests, documentation, and release evidence.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/product_requirements_acceptance_and_release_package/product_requirements_acceptance_and_release_package_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• VERIFY_ONLY
• APPLY_APPROVED

Tags and routing hints

• product
• operational
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Design human-in-the-loop workflows, triggers, states, exceptions, permissions, data contracts, idempotency, observability, rollback, and automation governance.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/business_process_automation_and_workflow_design/business_process_automation_and_workflow_design_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• VERIFY_ONLY
• APPLY_APPROVED

Tags and routing hints

• automation
• operational
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Define ownership, definitions, lineage, quality rules, access, retention, master data, issue management, and change control for decision-critical data.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/data_governance_quality_lineage_and_master_data/data_governance_quality_lineage_and_master_data_result.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• data_governance
• investigative
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Synthesize competitors, capabilities, moves, incentives, market signals, uncertainties, scenarios, indicators, and strategic options without unsupported prediction.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/competitive_intelligence_and_strategic_scenario_planning/competitive_intelligence_and_strategic_scenario_planning_result.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• intelligence
• investigative
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Create a fact-controlled crisis communication system with stakeholders, holding statements, approvals, channels, update cadence, rumor handling, accessibility, and post-event learning.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/crisis_communication_and_issues_management/crisis_communication_and_issues_management_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• VERIFY_ONLY
• APPLY_APPROVED

Tags and routing hints

• communications
• operational
• hybrid

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Develop truthful media narratives, briefing materials, Q&A, spokesperson preparation, bylines, and measurement while avoiding manufactured authority or deceptive influence.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/public_relations_media_briefing_and_thought_leadership/public_relations_media_briefing_and_thought_leadership_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• VERIFY_ONLY

Tags and routing hints

• communications
• operational
• hybrid

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Evaluate strategic fit, mutual value, roles, economics, governance, data, brand, risk, milestones, negotiation, and exit conditions for partnerships.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/partnership_alliance_and_ecosystem_strategy/partnership_alliance_and_ecosystem_strategy_result.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• partnerships
• investigative
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Produce concise decision-oriented reporting that reconciles metrics, highlights variance, risk, options, asks, ownership, and follow-through for senior stakeholders.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/board_executive_investor_and_management_reporting/board_executive_investor_and_management_reporting_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• VERIFY_ONLY
• APPLY_APPROVED

Tags and routing hints

• reporting
• operational
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Creates an approved project foundation, conventions, tooling, tests, documentation, and safe defaults from defined requirements and architecture.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/project_scaffolding_and_repository_bootstrap/project_scaffolding_and_repository_bootstrap_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• APPLY_APPROVED
• VERIFY_ONLY

Tags and routing hints

• engineering
• operational
• generation
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Turn meeting inputs into an agenda, decision log, commitments, owners, deadlines, risks, unresolved questions, and accountable follow-through without inventing consensus.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/meeting_intelligence_decision_capture_and_follow_through/meeting_intelligence_decision_capture_and_follow_through_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• VERIFY_ONLY

Tags and routing hints

• reporting
• operational
• hybrid

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Track policy and regulatory developments, stakeholder positions, scenarios, obligations, uncertainty, and response options using dated authoritative sources.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/policy_analysis_public_affairs_and_regulatory_monitoring/policy_analysis_public_affairs_and_regulatory_monitoring_result.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• public_affairs
• investigative
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Define material topics, boundaries, metrics, methods, controls, claims, targets, evidence, uncertainty, and stakeholder-ready impact reporting without greenwashing.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/sustainability_esg_and_impact_reporting/sustainability_esg_and_impact_reporting_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• VERIFY_ONLY
• APPLY_APPROVED

Tags and routing hints

• sustainability
• operational
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Assess affected people, rights, harms, benefits, distribution, contestability, oversight, redress, misuse, and residual risk for organizational or AI decisions.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/ethics_responsible_ai_human_rights_and_societal_impact/ethics_responsible_ai_human_rights_and_societal_impact_result.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• governance
• investigative
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Transform content for comprehension, accessibility, inclusive language, multiple reading levels, assistive technologies, and preserved legal or technical meaning.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/accessibility_plain_language_and_inclusive_content_transformation/accessibility_plain_language_and_inclusive_content_transformation_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• VERIFY_ONLY

Tags and routing hints

• accessibility
• operational
• hybrid

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Turn a task into a concise production-grade prompt with context, trusted inputs, variables, tool policy, output schema, quality gates, examples only when useful, and stop conditions.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/prompt_creation_and_specification/prompt_creation_and_specification_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• VERIFY_ONLY

Tags and routing hints

• prompt_engineering
• operational
• hybrid

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Improve an existing prompt for precision, efficiency, robustness, model fit, context economy, output reliability, and tool-aware execution without adding ornamental complexity.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/prompt_optimization_and_enhancement/prompt_optimization_and_enhancement_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• VERIFY_ONLY

Tags and routing hints

• prompt_engineering
• operational
• hybrid

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Evaluate prompts with fixtures, mutation tests, injection cases, schema checks, cross-model trials, failure analysis, and measurable acceptance criteria.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/prompt_review_evaluation_and_adversarial_testing/prompt_review_evaluation_and_adversarial_testing_result.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• prompt_engineering
• investigative
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Diagnose prompt failures, identify causal ambiguity or contract defects, apply the smallest repair, and verify improvement against unchanged fixtures.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/prompt_repair_refactoring_and_failure_resolution/prompt_repair_refactoring_and_failure_resolution_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• VERIFY_ONLY

Tags and routing hints

• prompt_engineering
• operational
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Design a report around audience decisions, questions, evidence, analysis, hierarchy, exhibits, uncertainty, recommendations, and action rather than document length.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/report_architecture_and_decision_narrative/report_architecture_and_decision_narrative_result.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• reporting
• investigative
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Produces accurate, audience-specific project status, public roadmap, release information, decisions, and risk communication from verified evidence.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/stakeholder_status_roadmap_and_release_communication/stakeholder_status_roadmap_and_release_communication_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• APPLY_APPROVED
• VERIFY_ONLY

Tags and routing hints

• enablement
• operational
• communication
• hybrid

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Produce a coherent report package with verified prose, tables, figure briefs, executive summary, appendices, accessibility, source traceability, and exact export review.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/report_production_visualization_brief_and_quality_assurance/report_production_visualization_brief_and_quality_assurance_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• VERIFY_ONLY

Tags and routing hints

• reporting
• operational
• hybrid

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Plan sponsorship, stakeholder impacts, readiness, communications, training, support, resistance, adoption metrics, reinforcement, and transition governance.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/change_management_adoption_and_enablement/change_management_adoption_and_enablement_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• VERIFY_ONLY
• APPLY_APPROVED

Tags and routing hints

• change_management
• operational
• hybrid

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Map stakeholders by interests, influence, impact, trust, information needs, participation, risks, and engagement strategy without manipulative targeting.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/stakeholder_mapping_engagement_and_coalition_planning/stakeholder_mapping_engagement_and_coalition_planning_result.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• change_management
• investigative
• hybrid

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Design roles, accountabilities, decision rights, forums, workflows, spans, interfaces, incentives, metrics, and escalation for a coherent operating model.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/organizational_design_operating_model_and_governance/organizational_design_operating_model_and_governance_result.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• organization
• investigative
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Assess quality objectives, process controls, procedures, records, deviations, corrective actions, training, audits, and improvement loops against defined criteria.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/quality_management_sop_audit_and_continuous_improvement/quality_management_sop_audit_and_continuous_improvement_result.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• quality
• investigative
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Design continuity and recovery plans, dependency maps, scenarios, exercise injects, decision logs, recovery tests, communication, and remediation tracking.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/business_continuity_disaster_recovery_and_tabletop_exercise/business_continuity_disaster_recovery_and_tabletop_exercise_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• VERIFY_ONLY
• APPLY_APPROVED

Tags and routing hints

• resilience
• operational
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Evaluate strategic position, portfolio choices, synergies, resource allocation, capital constraints, scenarios, risks, and measurable decision criteria.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/corporate_strategy_portfolio_and_capital_allocation/corporate_strategy_portfolio_and_capital_allocation_result.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• strategy
• investigative
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Create an evidence-driven innovation funnel with problem theses, idea diversity, stage gates, experiments, learning records, portfolio balance, funding, and kill criteria.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/innovation_pipeline_idea_portfolio_and_experiment_governance/innovation_pipeline_idea_portfolio_and_experiment_governance_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• VERIFY_ONLY

Tags and routing hints

• innovation
• operational
• hybrid

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Convert a real workflow into a model-ready task pack with context selection, retrieval boundaries, prompt variants, tools, schemas, evaluation fixtures, and operator guidance.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/language_model_task_intake_context_engineering_and_prompt_pack/language_model_task_intake_context_engineering_and_prompt_pack_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• VERIFY_ONLY

Tags and routing hints

• llm_operations
• operational
• hybrid

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Create a department plan connecting mandate, outcomes, portfolio, capacity, budget, dependencies, risks, operating cadence, metrics, and cross-functional agreements.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/department_operating_plan_and_cross_functional_coordination/department_operating_plan_and_cross_functional_coordination_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• VERIFY_ONLY
• APPLY_APPROVED

Tags and routing hints

• organization
• operational
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Makes the authoritative proceed, hold, or stop decision by evaluating completeness, evidence freshness, unresolved risk, rollback, operations, and communications.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• reports/project_health_completeness_and_release_readiness_gate/project_health_completeness_and_release_readiness_gate_decision.md
• typed_runtime_artifacts

Allowed modes

• VERIFY_ONLY

Tags and routing hints

• governance
• gate
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Design records classification, retention, legal holds, access, disposal, archival, provenance, search, and institutional-memory workflows.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/records_management_retention_and_institutional_memory/records_management_retention_and_institutional_memory_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• VERIFY_ONLY
• APPLY_APPROVED

Tags and routing hints

• knowledge_management
• operational
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Create truthful employee communications, listening loops, culture rituals, manager toolkits, feedback channels, accessibility, and trust-preserving engagement plans.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/employee_communication_culture_and_engagement/employee_communication_culture_and_engagement_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• VERIFY_ONLY

Tags and routing hints

• people
• operational
• hybrid

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Collect and analyze authorized threat information, indicators, tactics, infrastructure, campaigns, confidence, relevance, and defensive recommendations without enabling misuse.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/cyber_threat_intelligence_and_security_research/cyber_threat_intelligence_and_security_research_result.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• cyber_intelligence
• investigative
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Model fraud and abuse actors, incentives, attack paths, controls, signals, false positives, investigation workflows, privacy, and response options.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/fraud_abuse_and_integrity_risk_analysis/fraud_abuse_and_integrity_risk_analysis_result.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• risk
• investigative
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Define stages, data quality, pipeline coverage, forecast methodology, capacity, territories, quotas, scenarios, inspection, and management actions.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/sales_operations_pipeline_forecast_and_revenue_planning/sales_operations_pipeline_forecast_and_revenue_planning_result.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• sales_operations
• investigative
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Connect marketing, sales, onboarding, success, renewal, expansion, attribution, handoffs, data definitions, automation, and lifecycle reporting.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/revenue_operations_and_lifecycle_measurement/revenue_operations_and_lifecycle_measurement_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• VERIFY_ONLY
• APPLY_APPROVED

Tags and routing hints

• revenue_operations
• operational
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Design reliable administrative workflows for calendars, travel, correspondence, briefing, expenses, records, confidentiality, contingencies, and delegated authority.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/administrative_operations_travel_scheduling_and_executive_support/administrative_operations_travel_scheduling_and_executive_support_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• VERIFY_ONLY

Tags and routing hints

• administration
• operational
• hybrid

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Benchmark candidate language models against representative tasks, quality, safety, latency, cost, tool use, structured output, long context, and operational constraints.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/model_evaluation_benchmarking_and_selection/model_evaluation_benchmarking_and_selection_result.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• model_evaluation
• investigative
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Design language-model applications with retrieval, memory, tools, permissions, grounding, evaluation, observability, cost, failure handling, and human control.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/llm_application_rag_memory_and_tool_architecture/llm_application_rag_memory_and_tool_architecture_result.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• llm_engineering
• investigative
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Evaluate indexing, chunking, metadata, retrieval, ranking, citations, freshness, access control, answer faithfulness, latency, and failure cases.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/knowledge_retrieval_search_and_rag_quality_evaluation/knowledge_retrieval_search_and_rag_quality_evaluation_result.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• llm_engineering
• investigative
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Converts completed work, incidents, failures, and residuals into evidence-backed improvements without rewriting history or assigning blame.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• reports/retrospective_continuous_improvement_and_learning/retrospective_continuous_improvement_and_learning_investigation.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• governance
• investigative
• retrospective
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Make an independent deployment decision using use-case, model, data, safety, security, privacy, legal, monitoring, human oversight, rollback, and residual-risk evidence.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/ai_governance_model_risk_and_deployment_approval/ai_governance_model_risk_and_deployment_approval_result.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• VERIFY_ONLY

Tags and routing hints

• ai_governance
• gate
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Clarify only the unknowns that can change the selected graph, authority, evidence lane, output medium, budget, or acceptance decision.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/intent_interrogation_and_route_clarification/intent_interrogation_and_route_clarification_result.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• auto_orchestration
• investigative
• clarification
• hybrid
• auto_prompt

Do not use when

• the trigger condition is false
• another active capability already owns the complete outcome
• the added prompt would not change routing, safety, quality, or verification

When to use

Determine whether a skill is genuinely required, merely helpful, redundant with native prompts, or inappropriate for the task.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/skill_requirement_and_capability_fit_resolution/skill_requirement_and_capability_fit_resolution_result.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• auto_orchestration
• investigative
• routing
• factual
• auto_prompt

Do not use when

• the trigger condition is false
• another active capability already owns the complete outcome
• the added prompt would not change routing, safety, quality, or verification

When to use

Find and qualify exact skills only after a validated capability gap exists, preserving provenance, permissions, and fallback options.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/missing_skill_discovery_and_candidate_qualification/missing_skill_discovery_and_candidate_qualification_result.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• auto_orchestration
• investigative
• discovery
• factual
• auto_prompt

Do not use when

• the trigger condition is false
• another active capability already owns the complete outcome
• the added prompt would not change routing, safety, quality, or verification

When to use

Install one approved exact skill non-interactively into global .agents and OpenCode skill directories, verify both copies, and record provenance.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/governed_skill_installation_and_dual_location_registration/governed_skill_installation_and_dual_location_registration_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_APPROVED
• VERIFY_ONLY

Tags and routing hints

• auto_orchestration
• operational
• installation
• factual
• auto_prompt

Do not use when

• the trigger condition is false
• another active capability already owns the complete outcome
• the added prompt would not change routing, safety, quality, or verification

When to use

Create a narrowly scoped advanced skill only when discovery finds no suitable candidate, test it adversarially, and register verified copies for global and OpenCode use.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/skill_creation_evaluation_and_dual_location_registration/skill_creation_evaluation_and_dual_location_registration_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• APPLY_APPROVED
• VERIFY_ONLY

Tags and routing hints

• auto_orchestration
• operational
• skill_creation
• hybrid
• auto_prompt

Do not use when

• the trigger condition is false
• another active capability already owns the complete outcome
• the added prompt would not change routing, safety, quality, or verification

When to use

Execute any exact installed skill through a typed placeholder contract only when skill use is genuinely required and independently verifiable.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/generic_conditional_skill_execution_adapter/generic_conditional_skill_execution_adapter_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• APPLY_APPROVED
• VERIFY_ONLY

Tags and routing hints

• auto_orchestration
• operational
• skill_execution
• hybrid
• auto_prompt

Do not use when

• the trigger condition is false
• another active capability already owns the complete outcome
• the added prompt would not change routing, safety, quality, or verification

When to use

Loop any prompt, scenario, or exact skill only when repetition has measurable value, with separate batch and refinement loops, strict budgets, progress checks, and safe exits.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/bounded_prompt_and_skill_loop_orchestrator/bounded_prompt_and_skill_loop_orchestrator_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• PLAN_ONLY
• APPLY_SAFE
• APPLY_APPROVED
• VERIFY_ONLY

Tags and routing hints

• auto_orchestration
• operational
• loop_orchestration
• hybrid
• auto_prompt

Do not use when

• the trigger condition is false
• another active capability already owns the complete outcome
• the added prompt would not change routing, safety, quality, or verification

When to use

Independently determine whether a bounded loop should continue, complete, stop, escalate, or roll back using verified results, progress deltas, budgets, and safety state.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/loop_exit_plateau_and_completion_adjudication_gate/loop_exit_plateau_and_completion_adjudication_gate_result.md
• typed_runtime_artifacts

Allowed modes

• VERIFY_ONLY

Tags and routing hints

• auto_orchestration
• gate
• loop_gate
• factual
• auto_prompt

Do not use when

• the trigger condition is false
• another active capability already owns the complete outcome
• the added prompt would not change routing, safety, quality, or verification

When to use

Analyze, normalize, register, validate, and transactionally integrate one new Markdown prompt into every required Mission Directives routing, skill, template, fixture, catalog, evaluation, documentation, and integrity surface.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• MD-02
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/prompt-addition-registration-and-ecosystem-integration/prompt-addition-registration-and-ecosystem-integration_result.md
• prompt_addition_request
• prompt_addition_preview
• prompt_addition_receipt
• canonical_prompt_file
• updated_registry_and_evaluation_artifacts

Allowed modes

• PLAN_ONLY
• DRAFT_ONLY
• APPLY_SAFE
• APPLY_APPROVED
• VERIFY_ONLY

Tags and routing hints

• prompt_system
• operational
• ecosystem_mutation
• hybrid
• new-prompt
• add-prompt
• prompt-ingestion
• suite-maintenance

Do not use when

• another active capability already owns the complete outcome
• the requested result does not match this prompt's observable outcome
• required evidence or authority cannot be obtained safely

When to use

Reviews licensing, policy, records, ownership, approvals, retention, auditability, and applicable governance obligations.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• reports/license_compliance_and_governance_review/license_compliance_and_governance_review_investigation.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY
• VERIFY_ONLY

Tags and routing hints

• governance
• investigative
• review
• factual

Do not use when

• another active capability owns the complete requested outcome
• required evidence or authority is unavailable
• the task is a trivial transformation that does not need this capability

When to use

Execute repository mission promise deviation bloat and simplification audit under Mission Directives evidence, authority, template, skill, artifact, and verification contracts.

Inputs and prerequisites

• MD-00
• MD-01
• MD-03
• MD-04
• MD-02
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/repository-mission-promise-deviation-bloat-and-simplification-audit/repository-mission-promise-deviation-bloat-and-simplification-audit_result.md
• typed_runtime_artifacts

Allowed modes

• AUDIT_ONLY
• PLAN_ONLY

Tags and routing hints

• audit
• investigative
• analysis
• hybrid

Do not use when

• another active capability already owns the complete outcome
• the requested result does not match this prompt's observable outcome
• required evidence or authority cannot be obtained safely

When to use

Design or audit clinical operations as a safe workflow spanning intake, triage, scheduling, documentation, orders, results, follow-up, escalation, privacy, and accountable handoffs.

Inputs and prerequisites

• MD-00
• MD-01
• MD-02
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/healthcare-clinical-operations/healthcare-clinical-operations_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• APPLY_SAFE
• VERIFY_ONLY

Tags and routing hints

• healthcare_and_clinical
• operational
• operational
• hybrid

Do not use when

• another active capability already owns the complete outcome
• the requested result does not match this prompt's observable outcome
• required evidence or authority cannot be obtained safely

When to use

Conduct a privacy-conscious patient intake that captures identity, reason for care, history, medication/allergy safety, consent, accessibility, and urgent warning signs for qualified clinical review.

Inputs and prerequisites

• MD-00
• MD-01
• MD-02
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/patient-intake/patient-intake_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• APPLY_SAFE
• VERIFY_ONLY

Tags and routing hints

• healthcare_and_clinical
• operational
• operational
• hybrid

Do not use when

• another active capability already owns the complete outcome
• the requested result does not match this prompt's observable outcome
• required evidence or authority cannot be obtained safely

When to use

Create a patient-centered care-plan summary that accurately reflects clinician-approved diagnoses, goals, treatments, medications, monitoring, responsibilities, and escalation instructions in understandable language.

Inputs and prerequisites

• MD-00
• MD-01
• MD-02
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/care-plan-summary/care-plan-summary_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• APPLY_SAFE
• VERIFY_ONLY

Tags and routing hints

• healthcare_and_clinical
• operational
• operational
• hybrid

Do not use when

• another active capability already owns the complete outcome
• the requested result does not match this prompt's observable outcome
• required evidence or authority cannot be obtained safely

When to use

Quality-check a clinical note for identity, completeness, internal consistency, source attribution, medication/allergy safety, coding support, privacy, and required clinician correction without rewriting clinical judgment.

Inputs and prerequisites

• MD-00
• MD-01
• MD-02
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/clinical-note-qa/clinical-note-qa_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• APPLY_SAFE
• VERIFY_ONLY

Tags and routing hints

• healthcare_and_clinical
• operational
• operational
• hybrid

Do not use when

• another active capability already owns the complete outcome
• the requested result does not match this prompt's observable outcome
• required evidence or authority cannot be obtained safely

When to use

Assemble an insurance prior-authorization packet that matches payer criteria to clinician-supported evidence, required forms, coding, timelines, and submission/appeal tracking.

Inputs and prerequisites

• MD-00
• MD-01
• MD-02
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/insurance-prior-auth-packet/insurance-prior-auth-packet_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• APPLY_SAFE
• VERIFY_ONLY

Tags and routing hints

• healthcare_and_clinical
• operational
• operational
• hybrid

Do not use when

• another active capability already owns the complete outcome
• the requested result does not match this prompt's observable outcome
• required evidence or authority cannot be obtained safely

When to use

Triage a medical billing denial by verifying claim history, denial reason, coding/documentation support, payer rules, deadlines, and the correct correction, appeal, or contractual action.

Inputs and prerequisites

• MD-00
• MD-01
• MD-02
• MD-03
• MD-04
• runtime_context
• authorized_inputs
• project_evidence

Outputs and artifacts

• results/medical-billing-denial-triage/medical-billing-denial-triage_result.md
• typed_runtime_artifacts

Allowed modes

• DRAFT_ONLY
• APPLY_SAFE
• VERIFY_ONLY