Security audit
Security checks across malware telemetry and agentic risk
This skill edits local developer writing as advertised and includes review, stash, dry-run, and validation steps, with no evidence of hidden data access or exfiltration.
Install this only if you want an agent to rewrite files in the current repository. Run with --dry-run first, use a clean branch or committed worktree, and expect it may create a git stash, edit files across the codebase when --all is used, run validation tools, revert files that fail validation, and delete the review JSON after a successful run.
Partition remaining findings by `fix_safety`: **Safe Fixes** (auto-apply): - `chat_leak` - Delete conversational artifacts - `cutoff_disclaimer` - Delete knowledge cutoff references - `filler_phrase` - Delete filler phrases - `heading_restatement` - Delete restating first sentence
On successful completion (all validations pass): ```bash rm .beagle/ai-writing-review.json ``` If any validation fails, keep the file and report:
63/63 vendors flagged this skill as clean.
No suspicious patterns detected.
