Skip to content
cccskills
FreePublished on GitHub

agent-ops-constitution

Create/update .agent/constitution.md. Use when commands/boundaries/constraints must be confirmed before baseline or code changes. Draft v0 from repo evidence, then interview user.

The English summary is being prepared. Showing the original description.

See how to install

Files included (2)

  • SKILL.md6.1 KB
  • metadata.json1.5 KB

SKILL.md (original)

Review the instructions the agent will receive before you install.

Constitution workflow (mandatory before baseline)

Goal

Make .agent/constitution.md baseline-ready by confirming:

  • allowed/restricted/forbidden work areas
  • single-line build/lint/test/format commands
  • environment assumptions
  • project-specific constraints
  • issue-first workflow expectations

Pre-requisite: Context Map

Before starting constitution interview, ensure a context map exists:

  1. Check if .agent/map.md exists
  2. If NO map exists:
    📍 No context map found. Creating one first to inform constitution questions...
    
    → Invoke agent-ops-context-map to generate .agent/map.md
  3. If map exists but is stale (>30 days old or significant changes detected):
    📍 Context map may be outdated. Refresh? [Y]es / [N]o
    
  4. Use the map to:
    • Pre-populate candidate work areas
    • Identify build/test tooling
    • Understand project structure for interview questions

Procedure

  1. Ensure context map exists (see above)
  2. Run tool detection — invoke agent-ops-tools or aoc tools scan --save
    • Creates .agent/tools.json with available development tools
    • Identifies missing recommended tools based on detected project type
    • Populates "Available tools" section in constitution
  3. Inspect repo evidence (README, CI workflows, package/build files). Do not guess.
  4. Draft v0 constitution:
    • every inferred item must cite its evidence ("CANDIDATE from <path>")
    • anything without evidence is TODO + UNCONFIRMED
    • use detected tools to suggest build/lint/test commands
  5. Interview the user:
    • one question per TODO/UNCONFIRMED item
    • ask both "what should it be?" and "why?"
  6. Update constitution until:
    • build + test commands are CONFIRMED
    • work boundaries are CONFIRMED
  7. Use agent-ops-state to update .agent/focus.md.
  8. Invoke agent-ops-tasks for any setup work discovered.

Issue Discovery During Constitution

After constitution setup, invoke agent-ops-tasks discovery:

  1. Collect setup items discovered:

    • Missing configuration → CHORE issue
    • Broken scripts → BUG issue
    • Missing documentation → DOCS issue
    • Security concerns noted → SEC issue
    • Environment issues → CHORE issue
  2. Present to user:

    📋 Constitution setup found {N} items to address:
    
    High:
    - [BUG] Build script references missing file
    - [CHORE] CI workflow needs updating
    
    Medium:
    - [DOCS] README is outdated
    
    Create issues for these? [A]ll / [S]elect / [N]one
    
  3. Workflow continues:

    • Issues created before baseline capture
    • Baseline can now proceed with known issues documented

Template

Start from constitution template.

Low Confidence Completeness Check (MANDATORY)

For LOW confidence work, constitution must be verified as complete before proceeding.

Required Sections Checklist

When confidence is LOW, verify constitution has ALL of these sections CONFIRMED (not TODO/UNCONFIRMED):

SectionRequiredPurpose
Project scope✅What this project does
Work areas: allowed✅Where changes are permitted
Work areas: restricted✅Where caution is required
Work areas: forbidden✅Where changes are never allowed
Build command✅Single-line build command
Test command✅Single-line test command
Lint command✅Single-line lint command
Format command⚪ OptionalSingle-line format command
Coverage command✅ for LOWHow to measure coverage
Environment assumptions✅OS, runtime versions, dependencies
Tool detection✅.agent/tools.json exists
Constraints✅Project-specific limitations
Default confidence✅LOW/NORMAL/HIGH

Completeness Verification

Before LOW confidence work begins:

🎯 LOW CONFIDENCE CONSTITUTION CHECK

Checking constitution completeness for low confidence work...

| Section | Status |
|---------|--------|
| Project scope | ✅ CONFIRMED |
| Allowed areas | ✅ CONFIRMED |
| Restricted areas | ✅ CONFIRMED |
| Forbidden areas | ⚠️ TODO |
| Build command | ✅ CONFIRMED |
| Test command | ✅ CONFIRMED |
| Lint command | ✅ CONFIRMED |
| Coverage command | ❌ MISSING |
| Environment | ✅ CONFIRMED |
| Tool detection | ✅ .agent/tools.json exists |
| Constraints | ⚠️ UNCONFIRMED |

Result: INCOMPLETE — 3 items need confirmation

⚠️ Cannot proceed with low confidence work until constitution is complete.

Starting interview to fill gaps...

Staleness Check

For LOW confidence, also check constitution age:

AgeAction
< 7 daysProceed
7-30 daysPrompt: "Constitution is {N} days old. Review before proceeding?"
> 30 daysRequire: "Constitution is stale. Re-interview before low confidence work."

Gap-Filling Interview

If constitution is incomplete for LOW confidence:

  1. List missing/unconfirmed sections
  2. Invoke agent-ops-interview for each gap
  3. Update constitution with confirmed values
  4. Re-verify completeness
  5. Only proceed when all required sections are CONFIRMED
📋 Constitution gaps for low confidence work:

1. Forbidden work areas (currently: TODO)
2. Coverage command (currently: MISSING)
3. Constraints (currently: UNCONFIRMED)

Starting interview to fill these gaps...

Q1: What areas of the codebase should NEVER be modified?
(These will be marked as forbidden work areas)

Reviews

No reviews yet. Tell others what you thought.

Comments

No login needed. Share how it worked for you, ask a question, or note where it did not run.

No comments yet. Be the first to write one.

Comments are published right away. Do not include personal information such as email addresses or phone numbers. Up to 2 URLs.

0 / 1,000 characters

This form is protected by reCAPTCHA, and the Google Privacy Policy and Terms of Service apply.

More skills from this repository

[00] TnR ABSOLUTE RIGOR — Systemic Verification Doctrine. Forced cognitive exertion for high-stakes investigations. Mandates absolute deconstruction, mathematical proof chains, and ≥95% derived certainty. Zero-tolerance for heuristics or sentiment-based reasoning. Shortcuts are defined as intentional hallucinations.

The English summary is being prepared. Showing the original description.

diegosouzapw/awesome-omni-skill64Updated Mar 2, 2026

2agent

Free

Configures 2-Agent workflow between PM and implementation roles. Use when user mentions 2-Agent, 2エージェント, PM連携設定, Cursor設定, Cursor連携, 2-agent運用. Do NOT load for: 単独運用, ワークフロー実行, ハンドオフ処理.

The English summary is being prepared. Showing the original description.

diegosouzapw/awesome-omni-skill64Updated Mar 2, 2026

Appends relevant hashtags to journal entries based on content analysis. Intended for LLM execution.

The English summary is being prepared. Showing the original description.

diegosouzapw/awesome-omni-skill64Updated Mar 2, 2026

Add a new AI coding agent to Agent Sandbox. Creates all required files (Dockerfile, templates, CI, docs) and wires the agent into the CLI, proxy, and build system.

The English summary is being prepared. Showing the original description.

diegosouzapw/awesome-omni-skill64Updated Mar 2, 2026

Add a new property to the AI agents database. Use when the user wants to add, create, or introduce a new column, property, field, or feature to track across all agents in the comparison matrix. Handles all four required steps - database updates, groups.json, table display, and GitHub issue templates.

The English summary is being prepared. Showing the original description.

diegosouzapw/awesome-omni-skill64Updated Mar 2, 2026

Add new Agent templates to the meta-agent-skills framework.

The English summary is being prepared. Showing the original description.

diegosouzapw/awesome-omni-skill64Updated Mar 2, 2026

See all skills by diegosouzapw

Report a problem with this skill