plan-review

A PRD checker that certifies whether an existing product requirements document contains the information needed by later planning, coding, testing, and safety steps.

In plain words
What is it for?
Use it to check a PRD from a product perspective, an engineering perspective, or both.
Why use it?
It finds missing or unclear requirements before engineering work depends on them.

Skill for Claude CodeCodex

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add skills/ndisisnd/msg/plan-review
Any agent
npx skills add ndisisnd/msg --skill plan-review
Clone the repo
git clone --depth 1 https://github.com/ndisisnd/msg

Made for: Claude Code, Codex.

Per session 199 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,516 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00199 $0.01516
Opus 5 $0.00100 $0.00758
Sonnet 5 $0.00040 $0.00303
Haiku 4.5 $0.00020 $0.00152

Measured 2d ago against content hash bee2cf1822a7, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

plan-review scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

.claude/skills/plan-review/SKILL.md · 88 lines

How it starts

The opening of the file, as written. The whole thing — 88 lines — stays where its author put it; the contents beside it link to each section on GitHub.

plan-review

Usage

Invoke: /plan-review [prd-path] [--product | --eng]

  • Slash command: /plan-review
  • Natural language: "tune the PRD", "certify the PRD", "run the contract certifier", "check the PRD before plan-em"
  • Context: a path to an existing PRD .md file, or invocation immediately after plan-pm or plan-em saved one

Flags: --product (product tune) · --eng (eng tune) · neither → auto-selected in Step 1. refs/certification.md owns the checks, which tune runs which, the severity rubric, and the findings-table schema.

Path resolution and validation belong to script-cert-preflight.sh (Step 1) — this skill never re-derives them.

Posture

A certifier, not an adversarial reviewer: run the tune's check subset and nothing else — no check without a consumer. Never interview the user; a fix that needs a product decision is escalated as one batched pause, never a conversation.

Inputs

Name Format Required Source
PRD file path .md file path matching features/prd-*/prd-*.md Yes (asked if missing) User message, directory path, or description
Tune type flag --product or --eng No — auto-selected if missing (never asked) User message at invocation, or forwarded by plan-em
devkit/AHA.md Project learning log No — self-healing writeback skipped if absent Read + appended in Steps 1/3 if present
devkit/OPEN-QUESTIONS.md Ambiguity log No — deferred-decision log skipped if absent Appended in Step 3 if present
devkit/PLATFORMS.md Per-platform tolerance profiles No — check 6's bucket-coverage facet is skipped if absent Read by script-cert-mech.py (eng tune)

Outputs

Name Format Destination
Selected tune type Tune type: Product or Tune type: Eng inline End of Step 1
Certification findings Findings-table rows (refs/certification.md) <prd-dir>/reports/review-prd-[n]-[slug].md — one growing table; a pre-v5.4 PRD keeps its §7 Plan review findings section instead
Auto-fix terminal table # | Sev | Found | Fixed (refs/certification.md) Inline after fixes (Step 3)
Self-healing learnings Category-tagged [tune:<category>] entries Appended to devkit/AHA.md (Step 3), one per auto-fixed Critical/Major
Recurrence protocol-repair flag Inline warning when a category recurs across ≥3 runs Inline (Step 3)
Open questions table Normalized to # | Question | Answer | Status RESOLVED_PATH Open questions section (edited in place)
Revised PRD All Critical/Major (and selected Minor) fixes applied RESOLVED_PATH (edited in place)
Frontmatter stamp reviewed: yes (product-tuned: / eng-tuned: pair on a pre-v5.4 PRD) RESOLVED_PATH frontmatter

Read the full file on GitHub · 88 lines

Files

What ships with it

3 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 88 lines · 199 tokens per session scan A bee2cf1822a7

Subscribe to this mod's changes

plan-review is a skill published in the GitHub repository ndisisnd/msg (6 stars, last pushed 16d ago), licensed MIT. It adds 199 tokens to every session and 1,516 once invoked, about $0.0010 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.