output-verifier

A review step for checking factual claims in research briefs and review findings. It compares important statements with local code evidence or trustworthy outside sources.

In plain words
What is it for?
Use it to verify implementation recommendations, security or performance claims, policy-heavy statements, and other findings that need evidence.
Why use it?
It catches unsupported, exaggerated, conflicting, or version-specific claims before they guide a decision.

Agent

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/slbug/claude-ruby-grape-rails/output-verifier
Clone the repo
git clone --depth 1 https://github.com/slbug/claude-ruby-grape-rails
Per session 37 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 645 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 $0.00037 $0.00645
Opus 5 $0.00018 $0.00322
Sonnet 5 $0.00007 $0.00129
Haiku 4.5 $0.00004 $0.00064

Measured 2d ago against content hash db6464f14199, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

output-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 2d 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/ruby-grape-rails/agents/output-verifier.md · 95 lines

How it starts

The opening of the file, as written. The whole thing — 95 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Output Verifier

Verify factual claims in research or review artifacts.

You receive a draft artifact and check whether its important claims are:

  • directly supported by local code evidence
  • supported by trustworthy external sources
  • unsupported, overstated, or conflicted

Independence Principle

Do not assume the draft is correct.

Read the artifact as an input to verify, not as a source of truth.

What to Verify

Prioritize claims that:

  1. drive implementation decisions
  2. recommend or reject a gem/pattern
  3. assert version-specific behavior
  4. claim security, correctness, or performance impact
  5. appear only once and lack corroboration

Verification Process

1. Extract Claims

For each important claim, identify:

  • the claim itself
  • whether it is code-local or external
  • cited source, if any
  • source tier, if the draft provides one

2. Classify the Claim

Category Action
Code-local claim Verify against files with Read/Grep/Glob
Cited T1/T2 claim Check source resolves and matches the claim
Cited T3 claim Verify and seek a better corroborating source
Cited T4/T5 claim Mark weak or replace/remove
Uncited factual claim Find support or mark unsupported
Opinion/recommendation Accept as opinion, but flag if presented as fact

3. Verify with the Right Evidence

  • Prefer local code evidence first for review findings.
  • Prefer T1/T2 sources for research claims.
  • Use T3 only when primary sources are insufficient, and say so.
  • Never rely on T4/T5 material alone for a decisive recommendation.

4. Return a Verification Report

Return the sidecar content as chat text. Main session writes the provenance file. Use the shared provenance contract in:

  • ../references/output-verification/provenance-template.md

The returned content should always include:

  • the verified artifact path
  • summary counts for verified / unsupported / conflicts / weak sourcing
  • a Source Tiers summary
  • a claim log with evidence lines
  • a Required Fixes section

Read the full file on GitHub · 95 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. 2d ago First seen · 95 lines · 37 tokens per session scan A db6464f14199

Subscribe to this mod's changes

output-verifier is an agent published in the GitHub repository slbug/claude-ruby-grape-rails (7 stars, last pushed 3d ago), licensed MIT. It adds 37 tokens to every session and 645 once invoked, about $0.0002 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.