NOT RELEASED. No version bump - versions stay at 2.0.0 across all 9 plugins and
marketplace.json, because bumping is the release action and this ships only after
it has been tested. On a branch for the same reason.
WHY A SKILL AND NOT A FOURTH COMMAND. /security-audit-static is already mature -
sink analysis, self-refutation with attacker/victim rules, OWASP backstop, fan-out.
Rebuilding that inside something new would duplicate it. The hole in this plugin is
CORRECTNESS: there is no bug-finding review at all. So this is one skill with three
independently activated dimensions that defers to the existing command for security
and points at intended-vs-implemented for the doc-vs-code axis.
THE ANCHOR IS THE AGREEMENT, NOT THE FILE. The defects reviewers miss are rarely
visible inside one file - they are disagreements between two participants that each
read sensibly alone. Engine: map a flow, identify an obligation, inspect EVERY
participant, construct a violating execution, trace the consequence, refute, report.
Two lenses get a forced probe rather than a checklist mention: authority
reconciliation (a requested value is not an applied value) and identity correlation
(is the key unique, stable and live under overlap and reuse).
Refutation discipline is deliberately stricter than the security command's: a
correctness defect can harm only the person who triggered it and still be serious,
so the attacker/victim test does not transfer, and 'keep unless disproved' is too
permissive. Keep / Drop / Unresolved, with unresolved kept out of the findings list.
Parallelism fans out over complete flows, never over files - partitioning by file is
exactly the split that hides cross-boundary defects. Overlapping reads are allowed
and encouraged.
Coverage reports work performed in four states; zero findings is not 'not covered'.
Co-designed with GPT-6 Astra (Codex CLI). Contains no project-specific content: no
repo names, no paths, no bug identifiers, no defect text - verified by scan.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01G42vsxSKL7je39AsHZ5aJm
Codex CLI (0.136+) reads the same .claude-plugin/marketplace.json as Claude
Code, so PM Skills installs natively via `codex plugin marketplace add` — no
file-copying required.
- Add a first-class "Codex CLI (OpenAI)" install section (marketplace add +
per-plugin `codex plugin add`).
- Note the difference vs. Claude Code: skills work and are invocable by name;
`/slash` commands install but do not run as Codex slash commands.
- Document an optional path: ask Codex to convert the bundled command files
into skills (best-effort, model-driven).
- Remove the now-inaccurate "copy skills to .codex/skills/ — Skills only" row
from the "Other AI assistants" table.
Docs-only; no plugin/marketplace contents changed (counts unchanged, no
version bump). Validator: all 9 plugins pass.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
- Delete .docs/images/plugins-overview.webp (already replaced by plugins.png in v2.0.0).
- Delete .docs/images/how-skills-work.webp and its reference in README.md.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Introduce repo-level guidance for AI agents:
- CLAUDE.md: single source of truth (structure, design rules, versioning,
operational procedures). 65 skills, 36 commands, 8 plugins.
- AGENTS.md: thin pointer to CLAUDE.md for non-Claude agents.
- .gitignore: stop ignoring CLAUDE.md (now public); ignore CLAUDE.local.md
and keep ignoring _Internal/ (maintainer-only, never published).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Relocate image assets into .docs/images and update README image references accordingly. Four files were renamed/moved: add-marketplace.webp, how-skills-work.webp, plugins-overview.webp, and pm-skills-install.gif; README.md paths were updated to point to .docs/images/*. Keeps documentation assets organized under .docs/images.
Pawel HurynandClaude Opus 5
Fahri Meral
Admin