gjalla-prepare-and-commit
Pass
Audited by Gen Agent Trust Hub on Sep 6, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill requires the agent to ingest and review implemented code changes and specifications (Step 1), which creates a surface where malicious instructions in the code could influence the agent.
- Ingestion points: Implementation code, diffs, and the gjalla spec file (SKILL.md).
- Boundary markers: No explicit instructions to ignore embedded commands in the processed code or files are present.
- Capability inventory: git (add, commit) and the vendor's gjalla (ledger, loop, attest) CLI tools.
- Sanitization: No sanitization of the reviewed code or specifications is instructed.
- [COMMAND_EXECUTION]: The skill instructs the agent to execute various shell commands via
gitand the vendor'sgjallaCLI tools to manage the commit lifecycle and record verification evidence.
Audit Metadata