update-skill

update-skill is a command for Claude Code from kid-sid/claude-spellbook. It costs 0 tokens per session (625 once invoked), scanned A, original, MIT.

A review tool that reads pending findings from a JSON Lines file, where each line is a separate JSON record, and applies approved updates to skill files.

In plain words
What is it for?
Use it to list unreviewed findings, group them by skill, and update one selected skill at a time.
Why use it?
It gathers feedback in one place and helps turn reviewed findings into small, targeted documentation changes.

Command for Claude Code

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.

agentmods
npx agentmods add commands/kid-sid/claude-spellbook/update-skill
Clone the repo
git clone --depth 1 https://github.com/kid-sid/claude-spellbook

Made for: Claude Code.

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 update-skill

README.md
[![agentmods](https://agentmods.dev/badge/commands/kid-sid/claude-spellbook/update-skill.svg)](https://agentmods.dev/commands/kid-sid/claude-spellbook/update-skill)
Your own site
<a href="https://agentmods.dev/commands/kid-sid/claude-spellbook/update-skill"><img src="https://agentmods.dev/badge/commands/kid-sid/claude-spellbook/update-skill.svg" alt="Measured on agentmods" height="20"></a>
Per session 0 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 625 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
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 $0.00000 $0.00625
Opus 5 $0.00000 $0.00313
Sonnet 5 $0.00000 $0.00125
Haiku 4.5 $0.00000 $0.00063

Measured 5d ago against content hash 940b70357bac, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

update-skill 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 5d 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.

.claude/commands/update-skill.md · 44 lines

How it starts

The opening of the file, as written. The whole thing — 44 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Read .claude/findings.jsonl, show pending skill findings, and apply approved changes back to the relevant skill file.

Instructions

  1. Read .claude/findings.jsonl. If the file does not exist or has no entries where reviewed: false, say "No pending findings." and stop.

  2. Filter to reviewed: false entries. Group by skill field. Print a summary table:

| Skill         | Count | Sources           |
|---------------|-------|-------------------|
| fastapi       |   2   | web, gap          |
| (unmatched)   |   1   | web               |
  1. Ask the user which skill to process. Work on one skill per invocation.

  2. For entries where skill is null: list the query / url / note for each, suggest the most likely skill match by comparing keywords against folder names under skills/*/skill.md, and confirm with the user before proceeding.

  3. Read the target skills/<name>/skill.md.

  4. For each unreviewed finding for this skill, determine the minimal change:

    • source: "web" — new info from a search: add a checklist item, update a code example, or add a ### subsection if the topic is genuinely new
    • source: "gap" — pattern Claude had to invent: add a code example + checklist item in the most relevant existing section
    • source: "bug" — incorrect content: show the current block and the corrected version side by side as # BEFORE / # AFTER in a fenced block
  5. Show every proposed change in full before touching any file. Ask for explicit approval:

    • "Apply these changes? (yes / skip / edit)"
    • "skip" marks the finding reviewed without editing the skill
    • "edit" lets the user paste a corrected version
  6. On approval:

    • Edit skills/<name>/skill.md in-place — never rewrite the whole file, never change frontmatter or section order
    • For each processed finding, update its reviewed field to true in findings.jsonl (rewrite that line only)
  7. After editing, run /skill-validate to confirm the file still passes CI format checks. If it fails, show the error and offer to fix it before proceeding.

Read the full file on GitHub · 44 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. 5d ago First seen · 44 lines · 0 tokens per session scan A 940b70357bac

Subscribe to this mod's changes

update-skill is a command published in the GitHub repository kid-sid/claude-spellbook (187 stars, last pushed 1mo ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 625 tokens. 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.