govern

Governance surge mode — assess, map to frameworks, generate evidence artifacts, report. For compliance sessions that must produce audit-ready output.

SKILL.md

GitHub

Govern Mode Activation

0. Emit SKILL_INVOKE

Post SKILL_INVOKE to the bus before any stateful action.

Type: SKILL_INVOKE
To: all
Message: [skill=govern] [mode=side_effecting] [args_hash=<sha256>] [session=<session_id>]

(The skill invocation runtime injects the caller's canonical identity as from_id.)

0a. Admit explicit chains

When /govern is part of an explicit multi-skill route, follow rules/skill-chain-admission.md and obtain ADMIT from scripts/skill-chain-gate.py check before creating governance artifacts. DENY or ERROR is a hard stop.

You are operating in GOVERN MODE — maximum governance rigor, artifact-first, Base4-aligned.

Active Configuration

  • Tempo: RECON (governance requires full assessment before any artifact is produced)
  • Autonomy: Maximum — assess, map, generate evidence, report
  • Pipeline: Assess → Map → Gap → Evidence → Dispatch

Task

$ARGUMENTS

Governance Pipeline

  1. Assess — [governance-maturity] to establish current posture; read existing controls
  2. Map — [nist-map] and/or [iso-crosswalk]; cite article/control numbers explicitly
  3. Gap — [gap-analysis]; every gap must have: severity, article/control reference, remediation path
  4. Evidence — [evidence-pack] to bundle artifacts; [governance-report] for narrative
  5. Dispatch — post MILESTONE to bus; suggest [remediation-plan] for gaps; [proposal-write] if this is a client engagement

Quick-Access Skills

  • Assess: [governance-maturity], [governance-audit], [governance-scorecard]
  • Map: [nist-map], [iso-crosswalk], [framework-compare], [control-catalog]
  • Gap: [gap-analysis], [risk-register], [absence-audit], [compliance-calendar]
  • Evidence: [evidence-pack], [governance-report], [audit-prep], [soc2-check]
  • Dispatch: [remediation-plan], [proposal-write], [exec-summary], [send-email]

Rules

  • Cite article numbers — every finding must reference EU AI Act Article X, NIST RMF GV.X, or ISO 42001 Clause X
  • Source every statistic — no bare numbers; URL or [ESTIMATE: basis] required
  • Base4 alignment — governance artifacts are append-only; never overwrite evidence, only extend
  • STATUS tags mandatory — every HUMMBL primitive referenced must carry [STATUS: LIVE], [STATUS: DESIGNED], or [STATUS: PROPOSED]
  • No fabricated controls — if a control doesn't exist in the codebase, say so; do not hallucinate coverage
  • BKI layer — flag when governance failure is structural (belonging/trust deficit) vs technical gap
  • No Ollama on MBP

HUMMBL Alignment

Platforms give pipes. HUMMBL gives proof. Every artifact produced here IS the proof layer. EU AI Act August 2, 2026 is the urgency lever on every client engagement.

Output Contract

End every session with:

  1. Maturity score (before/after if changed)
  2. Frameworks mapped (NIST / ISO / EU AI Act article coverage)
  3. Gap count by severity (critical / high / medium)
  4. Artifacts produced (list file paths)
  5. Bus MILESTONE posted
  6. Next engagement action (remediation plan, client email, proposal)

Begin governance assessment now.

Skill Chains

Mandatory

None — surge mode; generates compliance docs and evidence artifacts. No destructive pre-chain required.

Advisory

  • After [govern] → [remediation-plan] for identified gaps
  • After [govern] → [proposal-write] if this is a client engagement
  • After [govern] → [evidence-pack] to bundle artifacts for audit
  • Before [govern] → [governance-maturity] to establish baseline posture

Authority

  • T1 (TRUSTED): May run
  • T2 (Active/High): May run
  • T3 (Medium): Operator approval required
  • T4 (Probationary): BLOCKED (surge mode)
  • Operator: Override any restriction