unattributed-claims-checker

unattributed-claims-checker is a skill for Claude Code from ur-grue/autopunk-media-skills. It costs 46 tokens per session (1,354 once invoked), scanned A, original, MIT.

A draft checker that identifies factual statements without a named source and adds an inline [NEEDS SOURCE] marker. It flags claims such as statistics, allegations, event descriptions, and expert assertions.

In plain words
What is it for?
Use it to audit articles, scripts, reports, and investigations, especially those involving serious allegations, statistics, or expert claims.
Why use it?
It shows editors exactly where attribution is missing before a draft reaches senior or legal review. This helps reduce unsupported claims in published or broadcast work.

Skill for Claude Code

Written for Claude Code: shipped in a Claude Code plugin.

Part of the autopunk-media-skills plugin — 187 skills shipped together

Good fit Use it to audit articles, scripts, reports, and investigations, especially those involving serious allegations, statistics, or expert claims.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/ur-grue/autopunk-media-skills/unattributed-claims-checker
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.

Any agent
npx skills add ur-grue/autopunk-media-skills --skill unattributed-claims-checker
Clone the repo
git clone --depth 1 https://github.com/ur-grue/autopunk-media-skills

Made for: Claude Code.

Or install autopunk-media-skills, the plugin that ships this one along with the rest of its 187 skills.

Wrote 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.

agentmods badge for unattributed-claims-checker

README.md
[![agentmods](https://agentmods.dev/badge/skills/ur-grue/autopunk-media-skills/unattributed-claims-checker/github.svg)](https://agentmods.dev/skills/ur-grue/autopunk-media-skills/unattributed-claims-checker)
Your own site
<a href="https://agentmods.dev/skills/ur-grue/autopunk-media-skills/unattributed-claims-checker"><img src="https://agentmods.dev/badge/skills/ur-grue/autopunk-media-skills/unattributed-claims-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.

agentmods 80×15 button for unattributed-claims-checker

Your own site · 80×15
<a href="https://agentmods.dev/skills/ur-grue/autopunk-media-skills/unattributed-claims-checker"><img src="https://agentmods.dev/badge/skills/ur-grue/autopunk-media-skills/unattributed-claims-checker.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 46 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,354 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe. Third-party audits
  • NVIDIA SkillSpector pass 7 Sept 2026
How audits are shown
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 5.1 $0.00046 $0.01354
Opus 5 $0.00023 $0.00677
Sonnet 5 $0.00009 $0.00271
Haiku 4.5 $0.00005 $0.00135

Measured 8d ago against content hash b101dbf2779a, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-12, from the pricing page.

Security

Grade A, and why

unattributed-claims-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 8d 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.

skills/editing/unattributed-claims-checker/SKILL.md · 83 lines

How it starts

The opening of the file, as written. The whole thing — 83 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Unattributed Claims Checker

What This Skill Does

Reads a draft article or script and flags every assertion of fact that lacks a named source, adding an inline [NEEDS SOURCE] marker so editors can see exactly where attribution is missing.

When To Use This Skill

  • Before you send a draft to a senior editor or legal review
  • When a piece has had multiple authors and sourcing discipline may be inconsistent
  • Any time you want a quick audit of a draft before publication or broadcast
  • When preparing an article that involves serious allegations, statistics, or expert claims

What You Need To Provide

Required: The full text of the draft — paste it directly into your message. Optional: A note about which claims are already known to have internal documentation (e.g., "the unemployment figures in paragraph 3 come from a government data release we have on file"). This prevents the assistant from flagging things you have already verified off-page.

How the Assistant Approaches This

  1. Reads the draft sentence by sentence and identifies every declarative claim — statistics, allegations, descriptions of events, characterisations of individuals or organisations, and predictions stated as fact.
  2. Checks whether each claim is clearly attributed to a named or sufficiently identified source (e.g., a person, institution, document, or data release) either within the sentence or in the immediately surrounding context.
  3. Appends [NEEDS SOURCE] directly after any claim that lacks attribution, then returns the full annotated draft so you can see every flag in context. It also produces a numbered list of flagged claims at the end, so you can work through them without scrolling the whole piece.

Output Format

  • The complete draft returned with [NEEDS SOURCE] inserted inline after each flagged claim — no other text is altered.
  • A numbered summary list beneath the draft: one line per flagged claim, quoting the flagged phrase and briefly explaining what type of attribution is missing (named source, official data, on-the-record quote, etc.).
  • Tone is neutral and editorial — no suggestions for how to find the source, no rewrites.
  • Length of output equals the draft length plus the summary list.

Read the full file on GitHub · 83 lines

Files

What ships with it

1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

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. 8d ago First seen · 83 lines · 46 tokens per session scan A b101dbf2779a

Subscribe to this mod's changes

unattributed-claims-checker is a skill published in the GitHub repository ur-grue/autopunk-media-skills (32 stars, last pushed 12d ago), licensed MIT. It adds 46 tokens to every session and 1,354 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 2026-09-04.