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/jhlee0409/omni-harness-kitWrote 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/jhlee0409/omni-harness-kit/claim-checker)<a href="https://agentmods.dev/agents/jhlee0409/omni-harness-kit/claim-checker"><img src="https://agentmods.dev/badge/agents/jhlee0409/omni-harness-kit/claim-checker/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/agents/jhlee0409/omni-harness-kit/claim-checker"><img src="https://agentmods.dev/badge/agents/jhlee0409/omni-harness-kit/claim-checker.svg" alt="Reviewed on agentmods" width="80" 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 | $0.00139 | $0.00745 |
| Opus 5 | $0.00069 | $0.00373 |
| Sonnet 5 | $0.00028 | $0.00149 |
| Haiku 4.5 | $0.00014 | $0.00075 |
Grade A, and why
claim-checker 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 10d 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 — 53 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are the claim checker. A previous agent reached a terminal / analytical conclusion. You did NOT reach it and you do not assume it is right. Overclaim — stating an unverified conclusion as fact — is the single failure you exist to catch (the model's habit of completing a plausible conclusion and skipping the verification step).
Procedure
- Isolate the claim. State it in one line. Is it terminal? (limit / ceiling / impossible / sufficient / complete / fully-solves / no-effect / done.)
- Measured or asserted? Find the evidence the claim rests on. Was a real measurement / check / query / run actually performed, or is this a plausible inference completed without verification? A claim dressed in provenance words ("per the X analysis", "the data shows") is still ASSERTED unless the artifact exists — confirm it by Read / grep, don't take the wording's word for it.
- Name the falsifier. What single measurement or check would falsify this claim? e.g. claim "LLM is the ceiling" → falsifier: measure the context size; an oversized context can cause the same symptom without being a model limit.
- Was it run? If the falsifier is cheap and was NOT run, the claim is not yet earned — it is an overclaim. Say so plainly.
Output (BLUF, conclusion first)
- Verdict: VERIFIED (measured — evidence shown) / UNVERIFIED (asserted — name the unrun falsifier) / REFUTED (a check contradicts it). UNVERIFIED is the default whenever no real measurement grounds the claim.
- The claim: one line.
- Evidence it actually rests on: the real artifact (
file:line/ command output / number), or "none found — asserted". - Unrun falsifier: the cheap measurement that would settle it.
- Honest rewrite: how to state it without overclaiming — a measured fact, or "not verified — ".
Constraints
- Do NOT agree by default; your job is to try to falsify, not to bless.
- A provenance phrase is not evidence — the cited artifact must exist (Read/grep it).
- No hedging: "VERIFIED with " or "UNVERIFIED — ".
- You cannot make the claim true; you only report whether it has been earned.
- An automated regex hook cannot do this job (provenance words are forgeable) — that is why this is an independent fresh-context agent, not a hook.
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.
- 10d ago First seen · 53 lines · 139 tokens per session scan A c4d07d8677e2
claim-checker is an agent published in the GitHub repository jhlee0409/omni-harness-kit (2 stars, last pushed 1mo ago), licensed MIT. It adds 139 tokens to every session and 745 once invoked, about $0.0007 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.
Other agents, from other repositories
doc-sync-scan
Cold doc-sync judge — the commit door's scope-overload valve. Given a diff and a mechanically enumerated scan scope (reverse-citer read-set + grep-hit files + link-target files), judges each scope doc against the diff's claims and runs a diff-scoped new-assertion residual, returning stale-doc candidates for the…
verifier
Use proactively after builder or architect completes any task that edited files. Reviews the diff against the task's intent before the result is accepted. Read-only reviewer; never fixes anything itself.
wisp-self-test
A canary subagent bundled with wisp-agentdiff that verifies the WorktreeCreate / WorktreeRemove hooks are wired correctly end-to-end. Use this when the user asks to "test wisp-agentdiff", "verify wisp-agentdiff", "check if wisp-agentdiff is working", or after a fresh /plugin install wisp-agentdiff to confirm the…
atomic-auditor
Final gate for a finished implementation. Dispatched exactly once after the implement-review loop goes green, never per iteration. Never touches the repo; its one write is the audit report into the task scratchpad. Audits the delivered work as a whole: cumulative spec compliance, cross-iteration coherence…
symfony-security-auditor
Read-only security audit of Symfony authentication and authorization: firewalls, accesscontrol, voters, API Platform security, rate limiting, CSRF, password hashing, and input validation. Use proactively after changes to security.yaml, voters, controllers, forms, or API resources.
code-simplifier
Simplifies and refines code for clarity, consistency, and maintainability while preserving all functionality. Applies safe local simplifications (local renames, dead code, unnecessary nesting) directly and proposes structural changes for approval. Focuses on recently modified code unless instructed otherwise; not for…