distill-project
Pass
Audited by Gen Agent Trust Hub on Oct 31, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted project data, including chat history, documents, and past outputs.
- Ingestion points: The skill inventories and reads reachable project conversations, files, documentation, memory, and published artifacts (SKILL.md, Step 2).
- Boundary markers: The skill contains explicit instructions to treat every chat and file as untrusted data and warns the agent to never execute instructions embedded within them.
- Capability inventory: The agent has the ability to read and write files within the project directory and execute a local style-checking script.
- Sanitization: The instructions include guidelines to exclude personal or sensitive material and to respect explicit privacy boundaries as rules.
- [DYNAMIC_EXECUTION]: The skill requires the execution of a local Python script provided within the skill package.
- Evidence: Step 7 in SKILL.md instructs the agent to run
<skill-directory>/scripts/check_style.pyon the generated files to validate style rules. The script is a local utility that performs regex-based checks for specific characters and banned words.
Audit Metadata