plan-reviewer

plan-reviewer is an agent for Claude Code from fulminate-io/knowledge-mcp. It costs 117 tokens per session (9,358 once invoked), scanned A, original, Apache-2.0.

A read-only skill for reviewing engineering plans as an adversarial senior engineer would. It checks the plan’s claims against the actual files, tests every stated criterion, and looks for missed cases or postponed problems.

In plain words
What is it for?
Use it to audit a proposed coding plan, verify cited files and requirements, test both successful and failing cases, and check whether a discovered problem was addressed across the whole project.
Why use it?
It catches incorrect assumptions and incomplete fixes before implementation, including claims such as “all” or “every” that lack complete evidence.

Agent for Claude Code

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 agents/fulminate-io/knowledge-mcp/plan-reviewer
Clone the repo
git clone --depth 1 https://github.com/fulminate-io/knowledge-mcp

Made for: Claude Code.

Wrote 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.

agentmods badge for plan-reviewer

README.md
[![agentmods](https://agentmods.dev/badge/agents/fulminate-io/knowledge-mcp/plan-reviewer.svg)](https://agentmods.dev/agents/fulminate-io/knowledge-mcp/plan-reviewer)
Your own site
<a href="https://agentmods.dev/agents/fulminate-io/knowledge-mcp/plan-reviewer"><img src="https://agentmods.dev/badge/agents/fulminate-io/knowledge-mcp/plan-reviewer.svg" alt="Measured on agentmods" height="20"></a>
Per session 117 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 9,358 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 5.1 $0.00117 $0.09358
Opus 5 $0.00059 $0.04679
Sonnet 5 $0.00023 $0.01872
Haiku 4.5 $0.00012 $0.00936

Measured 5d ago against content hash ae7eb65d99d7, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-05, from the pricing page.

Security

Grade A, and why

plan-reviewer 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 5d 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/agents/plan-reviewer.md · 612 lines

How it starts

The opening of the file, as written. The whole thing — 612 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Every thoughts(operation:"think") call passes origin:"plan-reviewer" (full stem, never reviewer).

A tool name written as thoughts(...) in this file is notation, not a literal tool id — in an MCP-prefixed environment call the prefixed form, e.g. mcp__knowledge__thoughts. When creating or rewriting a file, prefer Write/Edit over shell heredocs: the write tools are checked, quoted correctly, and leave a reviewable diff.

THE AUDIT LAWS

  1. VERIFY, DON'T TRUST. The plan is a signpost; every citation may itself have trusted a signpost. Open the cited files yourself.
  2. EXECUTE EVERY CRITERION. A criterion you did not run is a criterion you did not audit.
  3. AUDIT BOTH DIRECTIONS. Vacuous-pass AND fails-against-correct-work — only both catch all three defect families.
  4. COVERAGE CLAIMS CARRY EVIDENCE. "Every", "all", "nowhere", "the only" require an enumerated corpus. Absence findings are the dangerous kind.
  5. GENERALIZE EVERY FIX. One instance of a defect means sweep the whole set for its class before reporting.
  6. AUDIT PREMISE FIDELITY. Diff every rule the plan states against the ticket's original wording — a twisted premise makes the whole audit circular.

Read the full file on GitHub · 612 lines

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. 5d ago First seen · 612 lines · 117 tokens per session scan A ae7eb65d99d7

Subscribe to this mod's changes

plan-reviewer is an agent published in the GitHub repository fulminate-io/knowledge-mcp (1 stars, last pushed 2d ago), licensed Apache-2.0. It adds 117 tokens to every session and 9,358 once invoked, about $0.0006 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.