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.
npx skills add kyungseo/skillstead --skill docs-claim-checkgit clone --depth 1 https://github.com/kyungseo/skillsteadWrote 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.
[](https://agentmods.dev/skills/kyungseo/skillstead/docs-claim-check)<a href="https://agentmods.dev/skills/kyungseo/skillstead/docs-claim-check"><img src="https://agentmods.dev/badge/skills/kyungseo/skillstead/docs-claim-check/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.
<a href="https://agentmods.dev/skills/kyungseo/skillstead/docs-claim-check"><img src="https://agentmods.dev/badge/skills/kyungseo/skillstead/docs-claim-check.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 1 finding, up to medium
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- medium Excessive Agency · line 36 Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.Fix: Add human-in-the-loop confirmation for destructive, irreversible, or high-impact operations. Never auto-execute commands that modify files, send data, or alter system state.
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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00203 | $0.04256 |
| Opus 5.5 | $0.00081 | $0.01702 |
| Sonnet 5.5 | $0.00041 | $0.00851 |
| Haiku 4.5 | $0.00020 | $0.00426 |
Grade A, and why
docs-claim-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 13d 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.
How it starts
The opening of the file, as written. The whole thing — 348 lines — stays where its author put it; the contents beside it link to each section on GitHub.
docs-claim-check
Compare what public-facing documentation claims with what the provided evidence shows, one atomic claim at a time. Output findings with confidence labels and an explicit statement of what was (and was not) reviewed.
This skill renders a judgment backed by evidence. It is not a checklist, not a linter, and not a substitute for code review.
Before any tool use
Load this skill's instructions
Use the full SKILL.md already in context without reading it again. Otherwise load only this skill's SKILL.md at the path supplied by the host skill catalog, following the host's native skill/read or command-backed loading mechanism. Prefer native loading when the host offers a choice; another read tool's presence does not override the host's designated loader. This exception never applies to target or evidence reads.
Resolve an alias or relative path only from mappings or working-directory context
the host has already supplied. Do not guess or query missing values. For command-backed
loading, permit at most one attempt: a single cat -- '<literal SKILL.md path>'
command body (-- is optional when the path cannot be parsed as an option), safely
quoted without shell expansion. No wildcard, extra file, substitution, pipeline,
redirection, chaining, setup, or other command is allowed. These limits govern the
agent-supplied command body; they do not authorize adding a shell wrapper or changing
host startup behavior. Host permissions still apply; do not elevate or change settings.
A native loader may deliver the same file in chunks through its documented continuation mechanism; finish that loading operation before assessment. This does not authorize shell paging, discovery, or retrying a failed load. If the selected path cannot be resolved, loading is denied or fails, or its completed result is known to be partial/truncated, stop and request the full instructions or a compatible host. Do not proceed on a known incomplete body, repair the load with extra reads/commands, or switch loaders after failure. Use the supplied content and host completion/truncation signals; do not run a completeness probe. A permitted attempt that fails or is denied is a loading failure, not itself a boundary violation. A prohibited preparation attempt is a boundary violation even before assessment: stop and use the invalid assessment report below, whether or not the attempt executed.
What ships with it
4 files 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.
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.
- 13d ago Changed · +51 lines · +23 tokens per session ed75f896b1a4
- 1mo ago Changed · +54 lines 57da5087b98d
- 2mo ago Changed 3630b0833d64
- 2mo ago First seen · 243 lines · 180 tokens per session scan A 278a6ec3def3
docs-claim-check is a skill published in the GitHub repository kyungseo/skillstead (83 stars, last pushed 9d ago), licensed Apache-2.0. It adds 203 tokens to every session and 4,256 once invoked, about $0.0008 per session on Opus 5.5. 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 2026-08-30.
docs-claim-check compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| docs-claim-checkkyungseo/skillstead | Skill | 83 | 9d ago | Apache-2.0 | A | 203 | Claude Code, Codex |
| data-viz-rendererzebbern/claude-code-guide | Skill | 4.7k | 3d ago | MIT | A | 62 | Claude Code, Codex |
| reimagine-itKayforkind/reimagine-it | Skill | 212 | 5d ago | MIT | A | 230 | Claude Code |
| archforge-pptx-lintLove-Ash/archforge | Skill | 6 | 1mo ago | MIT | A | 91 | Claude Code |
| data-viz-genserejaris/kimi-skills | Skill | 11 | 3mo ago | MIT | A | 112 | Claude Code, Codex |
| data-viz-rendererserejaris/kimi-skills | Skill | 11 | 3mo ago | MIT | A | 62 | Claude Code, Codex |
| fireworks-tech-graphyizhiyanhua-ai/fireworks-tech-graph | Skill | 12k | 16d ago | MIT | A | 49 | Claude Code, Codex |