verify-claims

verify-claims is a skill for Claude Code from pandysp/claude-plugins. It costs 131 tokens per session (775 once invoked), scanned A, original, MIT.

A check that finds factual claims in an agent’s responses and looks for proof, especially for claims that support an important decision or recommendation.

In plain words
What is it for?
Use it before relying on claims in a design, commit, diagnosis, or user-facing recommendation.
Why use it?
It reduces the risk of treating a plausible but unverified explanation as a fact.

Skill for Claude Code

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

Part of the verify-claims plugin — 1 skill shipped together

Good fit Use it before relying on claims in a design, commit, diagnosis, or user-facing recommendation.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/pandysp/claude-plugins/verify-claims
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 pandysp/claude-plugins --skill verify-claims
Clone the repo
git clone --depth 1 https://github.com/pandysp/claude-plugins

Made for: Claude Code.

Or install verify-claims, the plugin that ships this one along with the rest of its 1 skill.

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 verify-claims

README.md
[![agentmods](https://agentmods.dev/badge/skills/pandysp/claude-plugins/verify-claims.svg)](https://agentmods.dev/skills/pandysp/claude-plugins/verify-claims)
Your own site
<a href="https://agentmods.dev/skills/pandysp/claude-plugins/verify-claims"><img src="https://agentmods.dev/badge/skills/pandysp/claude-plugins/verify-claims.svg" alt="Measured on agentmods" height="20"></a>
Per session 131 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 775 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.
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 $0.00131 $0.00775
Opus 5 $0.00066 $0.00387
Sonnet 5 $0.00026 $0.00155
Haiku 4.5 $0.00013 $0.00077

Measured 8d ago against content hash 86fcfb37eb29, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-08, from the pricing page.

Security

Grade A, and why

verify-claims 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.

plugins/verify-claims/skills/verify-claims/SKILL.md · 61 lines

How it starts

The opening of the file, as written. The whole thing — 61 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Verify claims: find the proof or say you can't

You construct narratives (plausible stories assembled from pattern-matching) and present them as conclusions. Sometimes confidently, sometimes hedged with "likely" or "probably." Either way, the user can't distinguish your verified knowledge from your confabulations without doing the verification work themselves. This skill forces that work onto you, where it belongs.

1. Identify claims

Scan your most recent substantive response for factual assertions:

  • Causal: "X caused Y", "this broke because…"
  • State: "the system does X", "the field contains Y", "the config is set to Z"
  • Historical: "this worked before", "it changed between A and B"
  • Attribution: "someone edited this", "the platform updated their policy"
  • Hedged: anything with "likely", "probably", "presumably", "appears to", "seems like", "suggests that." Hedging is the marker that you know you're uncertain but are stating the claim anyway.

2. Extract unverified claims

For each claim, ask: what is my proof for this specific claim? If the answer is any of these, the claim is unverified:

  • "My training data"
  • "It's the most plausible explanation"
  • "It's how these things usually work"
  • "I inferred it from context"
  • Nothing

List each unverified claim with: the exact text, what proof is missing, and whether it was stated confidently or hedged.

3. Plan verification

Pick the right method per claim type:

Claim type Verification method
Platform/API behavior Changelogs, docs, community posts
Current code/system state Read the code, grep the codebase
Historical state git log, git blame, execution logs
Causal ("X caused Y") Need proof for both: X happened AND Y follows from X
Quantitative Run the measurement
External fact Authoritative sources, expert communities

4. Execute and classify

Run each verification. Classify the result:

  • VERIFIED: definitive proof found. Cite it.
  • PARTIAL: some parts proven, others aren't. State exactly which.
  • INCONCLUSIVE: searched, found nothing either way. State what you searched. Do not restate the claim as a conclusion.
  • DISPROVEN: found contradicting proof. Correct the record.
  • UNVERIFIABLE: no feasible way to check. Explain why.

Read the full file on GitHub · 61 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. 8d ago First seen · 61 lines · 131 tokens per session scan A 86fcfb37eb29

Subscribe to this mod's changes

verify-claims is a skill published in the GitHub repository pandysp/claude-plugins (5 stars, last pushed today), licensed MIT. It adds 131 tokens to every session and 775 once invoked, about $0.0007 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.

Related

Other skills, from other repositories

systematic-debugging

Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.

obra/superpowers · 21 tokens

local-ai-agents

Build local-first AI agents that run entirely on a developer workstation with Microsoft Foundry Local and Qwen function-calling models. Covers Small Language Models (SLMs), the OpenAI-compatible local endpoint, sandboxed local tools, local RAG with Chroma, local MCP servers, hybrid cloud/local routing, and the…

microsoft/ai-agents-for-beginners · 200 tokens

next-cache-components-adoption

Turn on Cache Components in a Next.js app and resolve the blocking routes it surfaces. Use when the user wants to enable, adopt, or migrate to Cache Components, flip the cacheComponents flag, work through a flood of blocking-prerender / instant validation errors, run the cache-components-instant-false codemod, or…

vercel/next.js · 95 tokens

next-cache-components-optimizer

Drive a Next.js route to instant navigation by setting up an agentic loop, under Cache Components / PPR, on initial load (hard navigation) and client-side navigation (soft navigation). Encode the goal as a failing @next/playwright instant() e2e and work it to green, one verified route at a time; the shipped test then…

vercel/next.js · 170 tokens

next-partial-prefetching-adoption

Turn on Partial Prefetching in a Next.js app and work through the insights it surfaces. Use when the user wants to enable or adopt Partial Prefetching, flip the partialPrefetching flag, opt routes in with export const prefetch = 'partial', audit Link prefetch={true} behavior, preserve existing prefetched UI with…

vercel/next.js · 103 tokens

chronicle

Analyze Copilot session history for standup reports, usage tips, session search, and session reindexing. Use when the user asks for a standup, daily summary, usage tips, workflow recommendations, wants to search or find past sessions by keyword/file/PR, wants to reindex their session store, or asks about deleting…

microsoft/vscode · 72 tokens