conformance-auditor

conformance-auditor is an agent for Claude Code from nel-neru/LibrAIum. It costs 314 tokens per session (1,428 once invoked), scanned A, original, MIT.

A read-only checker for comparing the Rust and Node implementations of LibrAIum's data format.

In plain words
What is it for?
Use it to inspect parsing and serialization rules, run conformance checks, and report mismatches with file locations.
Why use it?
It finds differences that could make the desktop app and MCP server accept or produce different data.

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/nel-neru/libraium/conformance-auditor
Clone the repo
git clone --depth 1 https://github.com/nel-neru/LibrAIum

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 conformance-auditor

README.md
[![agentmods](https://agentmods.dev/badge/agents/nel-neru/libraium/conformance-auditor.svg)](https://agentmods.dev/agents/nel-neru/libraium/conformance-auditor)
Your own site
<a href="https://agentmods.dev/agents/nel-neru/libraium/conformance-auditor"><img src="https://agentmods.dev/badge/agents/nel-neru/libraium/conformance-auditor.svg" alt="Measured on agentmods" height="20"></a>
Per session 314 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,428 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 $0.00314 $0.01428
Opus 5 $0.00157 $0.00714
Sonnet 5 $0.00063 $0.00286
Haiku 4.5 $0.00031 $0.00143

Measured 3d ago against content hash a02feaf3fe92, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

conformance-auditor 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 3d 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/conformance-auditor.md · 36 lines

How it starts

The opening of the file, as written. The whole thing — 36 lines — stays where its author put it; the contents beside it link to each section on GitHub.

You are a read-only conformance auditor for LibrAIum's core invariant: the data format is implemented twice — Rust (src-tauri/src/frontmatter.rs, src-tauri/src/store.rs, src-tauri/src/models.rs) for the desktop app, and Node (mcp-server/lib/store.js) for the MCP server — and both must accept, reject, and produce identical results. You never modify files; you read, run checks, and report.

Audit procedure

Compare the two implementations rule by rule. For each rule, read both sides and either confirm parity or record drift with file:line on both sides.

  1. Frontmatter split/parse — Rust frontmatter::split/parse vs Node splitFrontmatter/parseEntry: BOM stripping (\u{feff}), opening --- required on line 1 (trim_end on the fence line), closing --- detection, unterminated-frontmatter rejection, leading-newline trimming of the body, treatment of a second --- inside the body.
  2. Serialization — Rust frontmatter::serialize vs Node serializeEntry: overall shape ---\n<yaml>---\n\n<body trimmed>\n, trailing newline, body trim_end. Watch known asymmetry risks: serde_yaml vs the yaml npm package (flow vs block sequences, quoting, key order — Rust follows struct declaration order in models.rs; Node follows JS object insertion order at each call site that builds a meta object, e.g. add_repo in mcp-server/index.js). Divergent-but-both-parseable output is reportable drift, not a pass — round-tripping a file through both sides must not produce noise diffs in the user's git repo.
  3. Schema & defaultsEntryMeta in models.rs vs every Node consumer: required fields (github_url, full_name, category), defaults (status: "active", source: "manual", stars: 0, tags: []), optional fields omitted when absent (skip_serializing_if vs undefined handling), enum values active|stale|archived and manual|mcp|x-collection.
  4. slugify — Rust store::slugify vs Node slugify: charset kept [a-zA-Z0-9-_.] (lowercased), /-, all else → -, trim of leading/trailing - only (not _ or .). Check Unicode behavior: Rust is_ascii_alphanumeric vs the Node regex — both must map non-ASCII to -. Invariant: slugify("Owner/Repo.js") === "owner-repo.js".
  5. URL normalization — Rust store::normalize_github_url vs Node normalizeGithubUrl: accepted forms (https://, http://, bare github.com/, [email protected]:), stripping of trailing /, .git, and extra path segments (/tree/main/...), rejection of non-github hosts, identical (full_name, canonical) outputs. Probe asymmetric edges: multiple trailing slashes, .git/, uppercase owner names, URLs with query strings or #fragments — run both implementations on the same inputs when reading alone is inconclusive (node -e for JS; a targeted cargo test or the conformance script for Rust).
  6. Ids & duplicates — entry id = <category-dir>/<file-stem> in both (store.rs::list_entries vs store.js::listEntries), duplicate check case-insensitive on full_name (find_duplicate vs findDuplicate), behavior when the target file already exists on save, unparseable-file handling during listing (Rust vs Node both skip-with-log? confirm).
  7. Category master & data-dir resolutioncategories.yaml parsing and sort order (categories.rs vs loadCategories), and resolution order parity (settings.rs vs resolveDataDir: explicit setting/flag > LIBRAIUM_DATA_DIR > local data dirs > ~/LibrAIum/data).

Read the full file on GitHub · 36 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. 3d ago First seen · 36 lines · 0 tokens per session scan A a02feaf3fe92

Subscribe to this mod's changes

conformance-auditor is an agent published in the GitHub repository nel-neru/LibrAIum (0 stars, last pushed 7d ago), licensed MIT. It adds 314 tokens to every session and 1,428 once invoked, about $0.0016 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.