Update DocsCAUTION

The AI toolkit for building reliable browser automations

The instruction file, verbatim from the audited commit — this is the text the model reads, and the surface the audit's instruction layer examines. Quoted here so you can judge it without cloning anything.

---
name: update-docs
description: Update AGENTS.md files based on session learnings about a topic. Use when documenting patterns, commands, or conventions discovered during development.
---

Update AGENTS.md files based on what was learned in this session about: $TOPIC

Read `./references/updating-agents-docs.md` for full guidelines before proceeding.

## Process

1. Review the conversation for learnings about the specified topic:
   - Patterns or conventions discovered
   - Commands needed but not documented
   - Architectural insights
   - Common pitfalls encountered

2. Determine where each learning belongs (lowest common ancestor):
   - Root `AGENTS.md` for repo-wide patterns
   - Package-level AGENTS.md for package-specific patterns
   - `docs/*.md` for detailed explanations

3. Check if the target AGENTS.md already has related content:
   - If multiple related items exist, consider extracting to a docs/ page
   - If adding would make the file too long (>60 lines for root), extract to docs/

4. Apply changes following the writing style in the guidelines

5. List what was updated and why

CAUTIONgrade B · trust 89/100 Install with care. The audit found things worth knowing before you trust its output.

What the source does

Filesystem

none-observed

Network

none-observed

Shell

none-observed

Dependencies

pinned

Secrets in source

none-found

Findings (5)

.claude/skills/glimpse-changes

.claude/skills/improve-codebase-architecture

.claude/skills/mintlify

.claude/skills/turborepo

.claude/skills/writing-tests

Gates applied: no_behavioural_pass.