autoskill
Pass
Audited by Gen Agent Trust Hub on Nov 10, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill analyzes session history and observation logs (untrusted data sources) to modify other skill files (SKILL.md), potentially allowing malicious instructions to be codified into the agent's permanent behavior.
- Ingestion points: Session history (user feedback, corrections, and repeated patterns) and external observation logs.
- Boundary markers: Mandatory summary presentation and explicit user approval are required before any file modifications are applied.
- Capability inventory: The skill has the capability to edit local files (SKILL.md) and execute git commands (git commit) across all active skill files.
- Sanitization: Includes a manual 'Signal quality filter' to verify if information is new, specific, and actionable, but it lacks technical sanitization or escaping of the ingested content.
Audit Metadata