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.
git clone --depth 1 https://github.com/andrei-isvoran96/claude-skill-labWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/commands/andrei-isvoran96/claude-skill-lab/plan-review)<a href="https://agentmods.dev/commands/andrei-isvoran96/claude-skill-lab/plan-review"><img src="https://agentmods.dev/badge/commands/andrei-isvoran96/claude-skill-lab/plan-review/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/commands/andrei-isvoran96/claude-skill-lab/plan-review"><img src="https://agentmods.dev/badge/commands/andrei-isvoran96/claude-skill-lab/plan-review.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00000 | $0.01013 |
| Opus 5 | $0.00000 | $0.00507 |
| Sonnet 5 | $0.00000 | $0.00203 |
| Haiku 4.5 | $0.00000 | $0.00101 |
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 9d 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.
How it starts
The opening of the file, as written. The whole thing — 72 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Draft an implementation plan for the user's request, then run up to 3 rounds of unbiased review by a fresh Plan subagent, revising between rounds.
The user's request: $ARGUMENTS
Procedure
Step 1 — Draft the initial plan
Spawn a Plan subagent to produce the first draft. Give it only the user's request above (no prior conversation). Ask for a step-by-step implementation plan: critical files, changes per file, architectural tradeoffs, test strategy, and any open questions.
Store the returned plan as PLAN_V1.
Step 2 — Round 1 review (always run)
Spawn a fresh Plan subagent as reviewer. It must not see your drafting reasoning or the parent conversation — only what you put in its prompt. CLAUDE.md is auto-loaded, so the reviewer already knows the architecture, commands, and coding standards.
The reviewer prompt must include:
-
The user's original request (verbatim:
$ARGUMENTS) -
The full plan text (
PLAN_V1) -
Instruction to verify claims against the real code — Read/Grep the codebase to check that referenced files, services, and patterns actually exist and are the right home for the changes. Do not judge on vibes.
-
Review rubric:
- Correctness: will this actually solve the user's request?
- Missing edge cases or error paths
- Scope: anything added beyond what was asked? Anything required but missing?
- Architectural fit: matches existing patterns?
- Testability and test coverage
- Adherence to the project's
CLAUDE.mdcoding standards
-
Required output contract — the reviewer MUST start its response with one of exactly two tokens on the first line:
APPROVED— no substantive issues, plan is ready to executeNEEDS_CHANGES— followed by itemized issues (severity: critical/warning/nitpick, with file:line references where applicable)
Nitpick-only findings still count as
APPROVED— only critical/warning issues warrantNEEDS_CHANGES.
Step 3 — Decide whether to iterate
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.
- 9d ago First seen · 72 lines · 0 tokens per session scan A d071d1361c7e
plan-review is a command published in the GitHub repository andrei-isvoran96/claude-skill-lab (1 stars, last pushed 3mo ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 1,013 tokens. 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.
Other commands, from other repositories
checklist
Generate a custom checklist for the current feature based on user requirements.
clarify
Identify underspecified areas in the current feature spec by asking up to 5 highly targeted clarification questions and encoding answers back into the spec.
specify
Create or update the feature specification from a natural language feature description.
analyze
Perform a non-destructive cross-artifact consistency and quality analysis across spec.md, plan.md, and tasks.md after task generation.
converge
Assess the current codebase against the feature's spec, plan, and tasks, then append any remaining unbuilt work as new tasks to tasks.md so implement can complete it.
implement
Execute the implementation plan by processing and executing all tasks defined in tasks.md.