verify
Pass
Audited by Gen Agent Trust Hub on Oct 24, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes Git commands (including status, rev-parse, add, rm, and write-tree) and arbitrary shell commands extracted from the project's own files, such as the README, CI workflows, and package manifests, to verify that work meets specific claims.- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content from various repository files to identify verification claims and the commands required to check them.
- Ingestion points: Project documentation (README), CI configuration files, package manifests, and project-specific plan or journal files (SKILL.md, references/report.md).
- Boundary markers: The instructions do not define delimiters or specific boundary markers for handling untrusted data ingested from the repository.
- Capability inventory: The skill can execute arbitrary shell commands found in the project files and perform Git repository operations.
- Sanitization: No evidence of sanitization, validation, or escaping of the extracted commands or plan data is provided before execution or processing.- [DYNAMIC_EXECUTION]: The skill utilizes a shell script recipe (detailed in references/tree-fingerprint.md) to dynamically compute a Git tree fingerprint by manipulating a temporary index file. It also performs dynamic execution of logic discovered within the project environment.
Audit Metadata