plan-generator
Pass
Audited by Gen Agent Trust Hub on Nov 6, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill generates plans containing shell commands intended for execution. These include repository management, environment setup, and testing commands such as
git log,git checkout,mkdir, andnpm test. - [EXTERNAL_DOWNLOADS]: The skill instructions mandate a final task in every plan that runs
npx ccusage@latest. This command fetches and executes the latest version of the ccusage utility from the npm registry to provide token usage reports. - [INDIRECT_PROMPT_INJECTION]: The skill generates executable tasks based on untrusted user requirements, which creates a potential surface for instructions within those requirements to influence the generated shell commands.
- Ingestion points: User requirements are parsed and analyzed in Step 1 of the execution process in
SKILL.md. - Boundary markers: The skill employs a structured markdown format for the resulting plans but does not explicitly instruct the agent to ignore or sanitize embedded instructions within the source requirements.
- Capability inventory: The skill utilizes
Bash,Write,Edit,Read, andSkilltools across its operations. - Sanitization: The instructions do not specify a process for sanitizing or validating requirements before they are used to populate the executable command fields in the plan.
Audit Metadata