claudemd-verifier

claudemd-verifier is an agent for coding agents from Ohswedd/praxis. It costs 66 tokens per session (391 once invoked), scanned A, original, MIT.

A read-only reviewer for proposed changes to CLAUDE.md files, which contain instructions loaded by Claude Code during work on a repository.

In plain words
What is it for?
Use it before changing a root or nested CLAUDE.md to decide whether the proposed version is safe or is a regression.
Why use it?
It checks that an edit does not remove valid guidance, working commands, or important project knowledge, and does not conflict with the code.

Agent

Part of the praxis plugin — 15 skills, 8 commands, 13 agents, 5 hooks shipped together

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/ohswedd/praxis/claudemd-verifier
Clone the repo
git clone --depth 1 https://github.com/Ohswedd/praxis

Or install praxis, the plugin that ships this one along with the rest of its 15 skills, 8 commands, 13 agents, 5 hooks.

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 claudemd-verifier

README.md
[![agentmods](https://agentmods.dev/badge/agents/ohswedd/praxis/claudemd-verifier.svg)](https://agentmods.dev/agents/ohswedd/praxis/claudemd-verifier)
Your own site
<a href="https://agentmods.dev/agents/ohswedd/praxis/claudemd-verifier"><img src="https://agentmods.dev/badge/agents/ohswedd/praxis/claudemd-verifier.svg" alt="Measured on agentmods" height="20"></a>
Per session 66 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 391 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 $0.00066 $0.00391
Opus 5 $0.00033 $0.00196
Sonnet 5 $0.00013 $0.00078
Haiku 4.5 $0.00007 $0.00039

Measured 5d ago against content hash 53f7f8550036, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

claudemd-verifier 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.

plugins/praxis/agents/claudemd-verifier.md · 37 lines

What it actually says

You protect project memory from silent degradation. A CLAUDE.md is loaded every session, so a bad edit poisons every future session. Read-only.

You are given the current CLAUDE.md and a proposed replacement (and optionally the structural diff from claudemd_check.py). Determine whether the proposal is a safe evolution or a regression.

Check that the proposal does NOT:

  1. Drop a still-valid instruction: any guidance, rule, or convention that is still true of the codebase.
  2. Remove a live command: a build/test/run/lint command that still works. Verify against the actual repo; a command may only be dropped if it genuinely no longer exists.
  3. Contradict itself or the code: introduce guidance that conflicts with another part of the file or with how the project actually works.
  4. Lose content to over-summarization: shrink by discarding signal rather than trimming redundancy.

For every removal or change the structural checker flagged, decide: is it justified (obsolete, contradicted by code, genuinely duplicated) or a regression? Require a concrete reason for each accepted removal.

Return SAFE or REGRESSION. If REGRESSION, list exactly what would be lost and what to restore. If SAFE, confirm each flagged removal is justified and why.

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 · 37 lines · 66 tokens per session scan A 53f7f8550036

Subscribe to this mod's changes

claudemd-verifier is an agent published in the GitHub repository Ohswedd/praxis (1 stars, last pushed 1mo ago), licensed MIT. It adds 66 tokens to every session and 391 once invoked, about $0.0003 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.