plannotator-setup-goal

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests user-supplied responses from a browser-based interview session and uses this data to guide codebase exploration and generate execution plans for future agent actions.\n
  • Ingestion points: Processes user data from goals/<slug>/interview-result.json and goals/<slug>/facts-result.json as described in Phases 2 and 3.\n
  • Boundary markers: The instructions do not define explicit delimiters or instructions to ignore embedded commands within the processed JSON files.\n
  • Capability inventory: The skill performs file system operations (mkdir), codebase exploration, and interacts with the plannotator CLI tool.\n
  • Sanitization: No explicit sanitization or validation of the user-provided text is performed before it influences the agent's reasoning.\n- [COMMAND_EXECUTION]: The skill relies on shell commands to manage project files and execute the goal setup toolchain.\n
  • Evidence: Uses mkdir -p goals/<slug>, plannotator setup-goal interview, and plannotator setup-goal facts to facilitate the goal discovery and planning workflow.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 07:58 AM
Security Audit — agent-trust-hub — plannotator-setup-goal