fact-check

A command that checks whether a document's claims match the actual codebase and corrects inaccurate statements in place.

In plain words
What is it for?
Use it to verify HTML, Markdown, or other text documents against source code and git history, and to add a verification summary.
Why use it?
It reduces the risk that documentation describes code, files, or history incorrectly.

Command

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 commands/theclaymethod/artifacture/fact-check
Clone the repo
git clone --depth 1 https://github.com/theclaymethod/artifacture
Per session 17 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,179 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin 88% copy Near-identical to another mod 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 $0.00017 $0.01179
Opus 5 $0.00009 $0.00589
Sonnet 5 $0.00003 $0.00236
Haiku 4.5 $0.00002 $0.00118

Measured 2d ago against content hash c9458faa4128, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

fact-check 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.

Origin

This is a copy

88% identical to fact-check — 10 lines differ, which has more behind it and is treated as the original. This page carries a canonical link to it rather than competing with it.

plugins/visual-explainer/commands/fact-check.md · 66 lines

How it starts

The opening of the file, as written. The whole thing — 66 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Load the visual-explainer skill, then verify the factual accuracy of a document that makes claims about a codebase. Read the file, extract every verifiable claim, check each against the actual code and git history, correct inaccuracies in place, and add a verification summary.

Source/artifact contract. If the target is generated HTML and a sibling or referenced .mdx/.tsx source exists, fact-check the generated artifact against reality but apply corrections to the source, then re-export the HTML with npm run ve:export. Edit generated HTML directly only when the artifact is legacy output or no source can be found; record that limitation in the verification summary.

Clarify. This is a Tier 2 command per ./references/clarify.md — it operates on an existing target and has no creative choices. Do not ask questions; just verify.

For HTML files: read ./references/css-patterns.md to match the existing page's styling when inserting the verification summary.

Target file — determine what to verify from $1:

  • Explicit path: verify that specific file (.html, .md, or any text document)
  • No argument: verify the most recently modified .html file in ~/.agent/diagrams/ (ls -t ~/.agent/diagrams/*.html | head -1)

Auto-detect the document type and adjust the verification strategy:

  • Generated HTML review pages (diff-review, plan-review, project-recap): detect from page content, locate MDX/TSX source when available, verify against the git ref or plan file the review was based on, correct source, and re-export
  • Plan/spec documents (markdown): verify file references, function/type names, behavior descriptions, and architecture claims against the current codebase
  • Any other document: extract and verify whatever factual claims about code it contains

Phase 1: Extract claims. Read the file. Extract every verifiable factual claim:

  • Quantitative: line counts, file counts, function counts, module counts, test counts, any numeric metrics
  • Naming: function names, type names, module names, file paths referenced in the document
  • Behavioral: descriptions of what code does, how things work, before/after comparisons
  • Structural: architecture claims, dependency relationships, import chains, module boundaries
  • Temporal: git history claims, commit attributions, timeline entries

Skip subjective analysis (opinions, design judgments, readability assessments) — these aren't verifiable facts.

Phase 2: Verify against source. For each extracted claim, go to the source:

  • Re-read every file referenced in the document — check function signatures, type definitions, behavior descriptions against the actual code
  • For claims about git history: re-run git commands (git diff --stat, git log, git diff --name-status, etc.) and compare output against the document's numbers
  • For diff-reviews: read both the ref version (git show <ref>:file) and working tree version to verify before/after claims aren't swapped or fabricated
  • For plan docs: verify that files, functions, and types the plan references actually exist and behave as described
  • For project-recaps: re-run git log commands to verify activity narrative and timeline

Classify each claim:

  • Confirmed: claim matches the code/output exactly
  • Corrected: claim was inaccurate — note what was wrong and what the correct value is
  • Unverifiable: claim can't be checked (e.g., references a file that doesn't exist, or a behavior that requires runtime testing)

Phase 3: Correct in place. Prefer source corrections. If the target is generated from MDX/TSX, edit the source and re-export the HTML artifact. If the target is markdown or legacy/no-source HTML, edit that file directly using surgical text replacements:

  • Fix incorrect numbers, function names, file paths, behavior descriptions
  • Fix before/after swaps (a common error class in review pages)
  • If a section is fundamentally wrong (not just a detail error), rewrite that section's content while preserving the surrounding structure
  • For HTML: preserve layout, CSS, animations, Mermaid diagrams (unless they contain factual errors in node labels or edge descriptions)
  • For markdown: preserve heading structure, formatting, and document organization

Read the full file on GitHub · 66 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 · 66 lines · 17 tokens per session scan A c9458faa4128

Subscribe to this mod's changes

fact-check is a command published in the GitHub repository theclaymethod/artifacture (2 stars, last pushed 9d ago), licensed MIT. It adds 17 tokens to every session and 1,179 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. It is 88% identical to fact-check, differing in 10 lines, and is treated as a copy.