planning-with-files

Pass

Audited by Gen Agent Trust Hub on Mar 10, 2026

Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection (Category 8) because it processes external data.
  • Ingestion points: WebFetch and WebSearch tools retrieve data that is stored in findings.md (SKILL.md).
  • Boundary markers: No markers or 'ignore' instructions are used to delimit external content in the provided templates (templates/findings.md).
  • Capability inventory: The agent has access to Bash and file writing tools (SKILL.md).
  • Sanitization: No sanitization of retrieved web content is performed.
  • [COMMAND_EXECUTION]: The skill uses hooks and instructions to run local scripts for session management, such as session-catchup.py and check-complete.sh. These are part of the skill's infrastructure.
  • [EXTERNAL_DOWNLOADS]: The skill utilizes WebSearch and WebFetch tools to retrieve information from the internet, which is then integrated into the agent's working memory via planning files.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 10, 2026, 04:40 PM