plannotator-setup-goal
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes the
plannotatorCLI tool via shell commands to process and gate generated documentation files (facts.md and plan.md). - [INDIRECT_PROMPT_INJECTION]: The skill ingests data from user input and local codebase files during the discovery and planning phases to generate project specifications. This creates a vulnerability where malicious instructions in the codebase could influence the output or the behavior of the planning tool.
- Ingestion points: User ideas and files explored in the local codebase as described in SKILL.md.
- Boundary markers: Absent. The skill provides no instructions to the agent to use delimiters or filter malicious content within the analyzed files.
- Capability inventory: Directory creation (
mkdir), file writing (facts.md, plan.md, goal.md), and tool execution (plannotator) as defined in SKILL.md. - Sanitization: Absent. There is no evidence of sanitization or safety checks performed on the codebase content before it is interpolated into the planning files.
Audit Metadata