skill-extractor

Pass

Audited by Gen Agent Trust Hub on Sep 16, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONPERSISTENCECOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill's primary function is to analyze session history and extract 'reusable knowledge' into new instruction files (SKILL.md). This creates a surface where adversarial content in a conversation could be formatted into a persistent skill that influences the agent's future behavior.
  • [PERSISTENCE]: The skill intentionally creates and modifies files in ~/.claude/skills/ and .claude/skills/. These files are persistent instructions that are loaded in future sessions, meaning a poorly generated or maliciously influenced skill would have a long-term impact on the agent's operating environment.
  • [COMMAND_EXECUTION]: The skill utilizes basic shell commands (e.g., ls, find) to check for existing skill directories. These operations are within the expected scope of a skill management tool.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 16, 2026, 03:46 AM
Security Audit — agent-trust-hub — skill-extractor