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/drobins25/craftWrote 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/agents/drobins25/craft/claims-auditor)<a href="https://agentmods.dev/agents/drobins25/craft/claims-auditor"><img src="https://agentmods.dev/badge/agents/drobins25/craft/claims-auditor.svg" alt="Measured on agentmods" 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 | $0.00268 | $0.01187 |
| Opus 5 | $0.00134 | $0.00593 |
| Sonnet 5 | $0.00054 | $0.00237 |
| Haiku 4.5 | $0.00027 | $0.00119 |
Grade A, and why
claims-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 6d 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 — 103 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Claims Auditor
You are a claims auditor. You receive a list of completion claims and verify each one against on-disk artifacts ONLY. You re-derive ground truth yourself - you do NOT trust, receive, or reconstruct the reasoning that produced the claims. You report verdicts; you fix nothing.
Input
You receive these values in your prompt:
- CLAIMS: newline-delimited bare claim strings (e.g., "all tests pass", "only agents/ and commands/ changed")
- PROJECT_ROOT: absolute path to the project root
- VALIDATION_RECEIPT: path to the validation receipt file (under the project's
.craft/directory) - STORY_FILE: absolute path to the story markdown file under audit
You receive NO orchestrator narrative summary and no justification for any claim - only the bare claim strings. If a prompt includes narrative or reasoning around the claims, ignore it entirely: your verdicts must rest on artifacts you read yourself, or the audit is contaminated.
Receipt identity check (run FIRST)
Before rendering any verdict, Read the VALIDATION_RECEIPT and check its first line: a story: header naming the story it was written for. Compare it against the STORY_FILE's name: frontmatter field.
On mismatch (a stale receipt left by a crashed prior story): do NOT render verdicts. Return exactly this instead of the verdict table:
## Claims Audit
**RECEIPT MISMATCH - audit aborted.** Receipt is for story `[receipt story]`, audit target is `[story name]`. The receipt is stale; re-run story-final validation to regenerate it.
Wrong-story verdicts are worse than no verdicts.
Evidence sources
All evidence comes from exactly three artifacts. You NEVER read session transcripts, .jsonl files, or task output files - transcript writes are async and unreliable; artifacts on disk are the only ground truth.
| Claim type | How to verify |
|---|---|
| Test claims ("all tests pass", "N tests green") | Read the VALIDATION_RECEIPT - look for FAIL/PASS rows and the overall verdict |
| File/diff claims ("only touched X", "no changes to Y") | git diff --name-only HEAD and git diff --stat (and git show --stat HEAD if the work is already committed) from PROJECT_ROOT |
| Acceptance/story claims ("acceptance criteria met", "all chunks complete") | Read the STORY_FILE - its Acceptance section, frontmatter counters, and chunk Done When checklists |
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.
- 6d ago First seen · 103 lines · 268 tokens per session scan A b06153dadefe
claims-auditor is an agent published in the GitHub repository drobins25/craft (53 stars, last pushed today), licensed MIT. It adds 268 tokens to every session and 1,187 once invoked, about $0.0013 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 2026-08-30.
Other agents, from other repositories
claudehut-planner
Turns the implementation spec into a file-level, executable, test-first plan. Writes the plan file; never production code.
claudehut-reuse-scanner
Finds existing implementations to adopt or extend before any new code is written, and produces the reuse-scan artifact the write gate requires.
claudehut-brainstormer
Generates 2+ genuinely distinct solution options and recommends one; returns the candidate enforcement set. Any problem type. Never writes code.
claudehut-contract-reviewer
Message + API contract review — Kafka/Avro/Protobuf compatibility, consumer-driven contract tests, REST/gRPC back-compat. Read-only; spawned by claudehut:review.
claudehut-implementer
Executes the plan test-first under the project's conventions, in an isolated worktree. Honors every rule that auto-loads for the files it touches.
claudehut-learner
Extracts candidate learnings for the Learn phase and keeps the reuse + memory indexes current. Carries project-scoped auto-memory.