process-directory-list

Pass

Audited by Gen Agent Trust Hub on May 7, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from a JSON file and interpolates it directly into prompts for sub-agents, creating a surface for indirect prompt injection attacks. \n- Ingestion points: The skill reads directory paths and item IDs from directory-instructions.json in the project root. \n- Boundary markers: Absent; instructions for sub-agents do not use delimiters (e.g., XML tags or triple backticks) to wrap the external data or include warnings to ignore instructions embedded within the directory names. \n- Capability inventory: The skill uses the Task tool to spawn sub-agents, invokes the /generate-instructions skill, and performs file system read/write operations on the configuration file. \n- Sanitization: Absent; there is no evidence of input validation, escaping, or filtering for the data read from the JSON file before it is used to construct prompts.
Audit Metadata
Risk Level
SAFE
Analyzed
May 7, 2026, 09:40 PM
Security Audit — agent-trust-hub — process-directory-list