CommunityCoding & Developmentgithub.com

josie2479-maker/claude-skills

Verify factual claims before delivering them. Use before any answer or deliverable the user will act on, forward, or share with others — rules, policy, legal, financial, date, or product facts — and whenever they say 'is that right', 'double-check that', 'verify this', 'fact-check', 'are you sure', or 'source?'. Exception: vetting a claim someone else handed the user is critical-thinking-walkthrough; fetching page text is reliable-fetch (this skill calls it).

What is claude-skills?

claude-skills is a Claude Code agent skill that verify factual claims before delivering them. Use before any answer or deliverable the user will act on, forward, or share with others — rules, policy, legal, financial, date, or product facts — and whenever they say 'is that right', 'double-check that', 'verify this', 'fact-check', 'are you sure', or 'source?'. Exception: vetting a claim someone else handed the user is critical-thinking-walkthrough; fetching page text is reliable-fetch (this skill calls it).

Works with✓Claude Code~Codex CLI~Cursor
npx skills add https://github.com/josie2479-maker/claude-skills/tree/HEAD/skills/fact-verification

Installed? Explore more Coding & Development skills: steipete/bluebubbles, steipete/eightctl, steipete/blucli · View all 6 →

Ask in your favorite AI

Open a new chat with this agent skill pre-loaded.

Documentation

Fact Verification

License: Released under CC BY 4.0 — free to use, adapt, and share with attribution.

1. Pull the claims

List every factual claim in the draft answer. Skip opinion and reasoning.

2. Triage

Mark each claim:

  • Structural: the user would decide, act, or forward based on it → must verify
  • Background: context only → label, don't chase

3. Verify load-bearing claims

  • Check against a primary source (the publisher's own site, the statute or standard itself, official documentation), not a summary

  • Use reliable-fetch for page text, if you have it; otherwise pull the page with a real browser tool rather than a summarizing fetch. If a source can't be reached, say so; don't fill in from memory

  • Dates, deadlines, amounts: confirm against the source's own wording, and note its as-of date

  • Forecasts and warnings: a forecast whose time has passed is not a result. Label every number as observed or forecast. For a past forecast, peak or alert level, look up the observed record (the measurement log or the system's recorded data), not the forecast or alert text, and report the actual. A number from a withdrawn forecast never happened.

  • Find the exact figure or wording in the source itself. A working link, a real journal and a matching topic don't prove the source says it. Several sources agreeing counts only if they didn't copy each other.

  • Citations carried in templates or older documents can be out of date. Check each cited rule, clause or section against the current official text, not the template.

  • A negative ("no requirement", "not found") covers only the sources and words actually searched. Before stating one:

    • search every layer that can create the duty (law, regulation, official guidance, local rules) and the project's own records;
    • search the concept with synonyms, not one keyword;
    • scope the claim to what was searched ("no regulation found in the sources checked", not "no requirement").

    Keep that scope in the headline itself. Caveats placed below a headline don't travel with it when it's copied.

  • Separate what a rule permits from what happened in the user's own case. Requests, letters, approvals and decisions come from the user's records (their notes, decision and event logs, earlier research in the project), not from the rule itself. Search those first, and ask where the record lives if the obvious place is empty.

  • Product plans ("is X planned?"): check the vendor's help-center "current limitations" sections first, then release notes, then blogs and press.

  • An arithmetic discrepancy handed to you: re-add every level (line items, subtotals, totals) from the source. The given explanation may cover only one of several gaps.

4. Label in the answer

  • ✅ Verified: [source, date checked]
  • ⚠️ Unverified: from training data, may be stale
  • ❓ Guess/inference: say what it rests on

5. Before delivering

Re-read steps 2–4. Any load-bearing claim without ✅ or an explicit ⚠️ → fix it or flag it.

Individual skills in this repo

This repo contains 3 individual skills — each has its own dedicated page.

josie2479-maker/claude-skills

Test a ready-made list of deliverables, documents or artifacts before anyone commits to tracking it. Use when someone is handed a list from a template, vendor, consultant, PMO framework or AI (\"standard project documents\", \"required artifacts\", \"deliverables checklist\") and asks whether to adopt it, what is really required, what to drop, or what is missing. Tags each item by who requires it (funder, law or regulation with citation, a named person, or standard practice that nobody requires), reads the funder's or agency's own template section by section to find required items the list leaves out, and returns a table plus a drop / keep / add list. Not for building a list from scratch with no list in hand, and not for checking whether finished documents are good.

josie2479-maker/claude-skills

Interview the user about something they are ABOUT TO BUILD, before any of it is built — a feature, a system, an automation, a workflow, a document structure. Use when a plan exists only in their head or as a rough idea and the design decisions are still open: \"grill me\", \"interview me about this\", \"help me think this through before I build it\", \"what am I not deciding\", \"ask me what you need to know\". Works the design as a dependency tree, asking one answerable question per round (the one that unblocks the most decisions first) with a recommended answer attached, and never asking anything it could look up itself. BOUNDARY — this is front-loaded, for decisions not yet made. Reviewing a plan that already exists and stress-testing it for holes is the plc skill. Interrogating information someone else handed you is critical-thinking-walkthrough. If the thing is already written down and the question is \"is this any good\", it is not this skill.

josie2479-maker/claude-skills

Strip AI-essay voice from a document that will be shared outside this conversation — a memo, spec, brief, proposal, SOW, report, or any deliverable going to a supervisor, agency, client, or contractor. Use when asked to \"clean this up\", \"make this professional\", \"take the Claude out of it\", \"de-AI this\", \"make it ready to send\", or when a draft is about to be handed off and has not had a register pass. Also run it unprompted as the last step when drafting any external-facing document. Not for chat replies, personal notes, or internal scratch files — conversational voice is fine there and flattening it is a loss. Not for fiction or other creative writing, where the voice is the point; see the fiction note in the body.

Related Skills