Skip to content
Back to skills

Agentdoctor Release Freeze

ASecurity

Pre-freeze / post-release integrity audit for AgentDoctor 3.0.0. Use when verifying freeze readiness, auditing implementation integrity, or checking whether remediation may be committed—never for inventing features or performing release actions.

  • 10 stars
  • 0 votes
  • 0 copies
  • 0 views
  • Added September 28, 2026
ai-agentsgogitsecurity

Works with

  • cli
  • mcp

Security analysis

A100/100

Scanned September 28, 2026

npx -y skills add pranee54/AgentDoctor --skill agentdoctor-release-freeze --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Agentdoctor Release Freeze?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Agentdoctor Release Freeze
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/pranee54-agentdoctor-release-freeze/badge)](https://www.skillsdirectory.com/skills/pranee54-agentdoctor-release-freeze)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: agentdoctor-release-freeze
description: Pre-freeze / post-release integrity audit for AgentDoctor 3.0.0. Use when verifying freeze readiness, auditing implementation integrity, or checking whether remediation may be committed—never for inventing features or performing release actions.
---

# AgentDoctor release freeze

## Purpose
Prove the **current 3.0.0** tree is correct, secure, and honestly documented before leaving it untouched. Version is already released.

## When to use
- Final freeze / integrity audits
- After security remediation on a dirty tree
- Before human commit of remediation (do not auto-commit)

## Non-negotiable rules
- DO NOT bump version, publish, tag, push, or create GitHub releases
- DO NOT invent features or redesign architecture
- DO NOT weaken security or delete tests to go green
- PASS requires evidence; NOT_VERIFIED must not become PASS
- Ownership ≠ containment (`OWNERSHIP_BOUNDARY_VERSION`)

## Workflow
1. Read `docs/internal/AGENTDOCTOR_3_0_CURRENT_TRUTH.md` (authoritative pointer)
2. Inventory CLI / dashboard / MCP from **source**, not memory
3. Re-run hostile suites under `tests/unit/project/` + MCP transport + provider E2E
4. Fix only proven defects; add regressions
5. `npm run verify` then `npm pack` + clean-install smoke
6. Update freeze audit docs; leave commit to human

## Verification checklist
- [ ] `npm pkg get version` = 3.0.0 and `PACKAGE_VERSION` matches
- [ ] CLI `--version` = 3.0.0; dashboard HTML / `ops.version` = 3.0.0
- [ ] OPEN_P0=0 OPEN_P1=0 for security/correctness
- [ ] RELEASE_ACTIONS_PERFORMED=NONE
- [ ] Dirty tree reported if remediation present

## Failure conditions
- Any foreign/`.private`/symlink path becomes VERIFIED project intelligence
- Stale Brain/Twin/Graph treated as current truth
- Model/dashboard/MCP grant write without real approval
- Hang on `$HOME` / broad roots

## What NOT to do
Roadmaps, provider additions, version bumps, silent doc rewrites that erase historical audits.

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…