Fact Checker
Verifies factual claims in your documents against official sources and applies corrections only after you approve them.
Content & WritingBeginner★ 1,323⑂ 212AI score 9/10Last updated: Aug 8, 2026
What it does
- Scans a document for verifiable claims — specs, version numbers, release dates, statistics, benchmark scores, pricing — while skipping opinions, prose, and tutorial steps.
- Searches authoritative sources for each claim: official product pages, API docs, release notes, GitHub releases, and package registries (npm, PyPI, crates.io).
- Produces a comparison table and fact-check report with statuses (✅ accurate, ❌ incorrect, ⚠️ outdated, ❓ unverifiable), including line locations, current text, proposed fix, source link, and rationale.
- Never edits silently: it waits for explicit approval, then applies edits, verifies them, and reminds you to commit.
- Adds good hygiene: temporal qualifiers ("as of January 2026"), numeric precision matched to the source, and inline citations.
Who it's for
- Technical writers, bloggers, and book authors whose docs contain fast-moving facts (AI model specs, library versions).
- Engineering teams refreshing stale wikis, READMEs, or onboarding guides.
- Analysts and marketers who want sourced numbers before a report or proposal goes out.
Examples
- Model spec refresh: "Fact-check the context windows in section 2.1" → outdated model names and token counts are tabled against current official docs, then fixed on approval.
- Benchmark verification: "Verify the benchmark scores in chapter 5" → each number compared to the official publication, with only the mismatches flagged and linked.
- Dependency audit: "Are the library versions in the README still current?" → registry lookups surface outdated pins with changelog references for the update.
· · · Install guide · · ·
Install in the Claude app (no terminal)
- Download the ZIP with the button below.
- In Claude, open Settings → Capabilities and turn on 'Code execution and file creation'. (one time)
- Go to Customize → Skills → + → 'Upload a skill' and upload the ZIP.
Install in Claude Code
Let Claude do it — paste this into Claude Code
Install the skill I found on Claude Skill Mart. Copy the fact-checker folder from the GitHub repo daymade/claude-code-skills into my ~/.claude/skills/fact-checker/. When it's done, tell me in one line what this skill can do.
Install with a command instead
git clone https://github.com/daymade/claude-code-skills.git && mkdir -p ~/.claude/skills && cp -r claude-code-skills/fact-checker ~/.claude/skills/⚠ This is a third-party skill. Check the source repository before installing.
- Open a terminal (Terminal on macOS/Linux, Git Bash or WSL on Windows).
- Clone the skills repository:
git clone https://github.com/daymade/claude-code-skills.git - Make sure the Claude Code skills directory exists:
mkdir -p ~/.claude/skills - Copy just this skill into place:
cp -r claude-code-skills/fact-checker ~/.claude/skills/ - Confirm the install:
ls ~/.claude/skills/fact-checkershould showSKILL.md. - Restart Claude Code and allow web tools (WebSearch/WebFetch) — the skill depends on live source lookups.
- Open the file you want checked and ask: "Fact-check this document" or "Verify the numbers in docs/guide.md".
- Review the report; the skill only rewrites your file after you say "apply the corrections". Commit to git first for a safe rollback.
View source on GitHub ↗License: MIT