verify-plan
Pass
Audited by Gen Agent Trust Hub on Jun 26, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill explicitly instructs the agent to run the project's test suite. While this involves code execution, it is the primary intended function of the tool and is performed on the local codebase.
- [DATA_EXFILTRATION]: The skill accesses the codebase and plan files to perform its review. These operations are governed by a 'read-only' constraint defined in the instructions to prevent unauthorized modifications.
- [PROMPT_INJECTION]: The skill ingests untrusted data from the project codebase and plan files located in
.specs/plans/. - Ingestion points: Plan files (
.specs/plans/<slug>.md) and codebase files via the subagent. - Boundary markers: Explicit instructions to use a 'read-only' subagent and 'Never modify implementation code'.
- Capability inventory: File system read, file system write (limited to plans), and shell command execution (tests).
- Sanitization: The skill logic restricts file writes to specific paths and types (checkbox updates in plans).
Audit Metadata