Honesty Auditor — Development agent for Claude Code
Reviews a change for the product-critical honesty rules — provenance, confidence, layer labels, scope denominators, error bars, and the refusals.
How to install Honesty Auditor
Installs to ~/.claude/agents/assaio-assaio-honesty-auditor.md
mkdir -p ~/.claude/agents && curl -fsSL https://raw.githubusercontent.com/assaio/assaio/HEAD/.claude/agents/honesty-auditor.md -o ~/.claude/agents/assaio-assaio-honesty-auditor.md Restart Claude Code, or start a new session, for it to be picked up.
What Honesty Auditor does
name: honesty-auditor description: Reviews a change for the product-critical honesty rules — provenance, confidence, layer labels, scope denominators, error bars, and the refusals. Use on any change that adds, renames, or reshapes a number a user reads. Read-only. tools: Read, Grep, Glob
You audit one thing: whether a change can make `assaio` state something it has not earned. That is the only failure this project cannot absorb — a wrong number here is indistinguishable from a right one
Alternatives in Development
- Sanitizer — Generic sanitizer worker agent 262 ★
- Ontology Term Lookup — Use this agent when you need to find ontology terms by their textual labels or descriptions using the OLS4 MCP 163 ★
- Probability Honesty Checker — Forces honest first-principles probability distribution + EV calculation 139 ★
Full documentation available on GitHub
View Source RepositoryRelated Agents
Bk Researcher
Answers a technical question a skill cannot verify locally, from independent sources and the pinned major's do
Lesson Writer
Drafts the Hebrew lesson for a completed pipeline step from the step brief, report, diff and review — facts an
Swagger Documenter
Enriches the FastAPI OpenAPI/Swagger docs of the BookShelf backend. Adds summary, description, response_descri
Cr Silent Failure Hunter
Hunts silent failures in the PR diff — empty or overly broad catch blocks, fallbacks that mask errors, excepti
Golang Reviewer
Fast Go code review focused on idioms, error handling, concurrency bugs, and performance issues. Returns only
Review Bug
Review: Bug/Logic Error Detection (coverage-first) — report all findings with confidence and severity tags; do
Related Skills
Paper Reader
A Claude Code skill for critical paper reading: 4 modes (deep-read, quick-scan, recommend, lit-review) with ev
OSS Migration Eval
Decide whether to switch an LLM pipeline to a cheaper/open-source model — honest cost-vs-accuracy verdict with
Honesty
Reviews your own report before delivering — separates what was verified from what was assumed, and declares th