implement-spec
Pass
Audited by Gen Agent Trust Hub on Oct 27, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructions explicitly require the agent to read the repository's CI entry point (such as a Makefile or CI workflow) and execute each of its steps locally (Step 8). This results in the execution of arbitrary shell commands defined within the repository's configuration.
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and act upon external data from specifications and issue tracker tickets, creating an attack surface where malicious instructions in these sources could influence the agent's behavior.
- Ingestion points: Specification files, task tickets (Step 1), and repository CI configuration files (Step 8).
- Boundary markers: Absent; the skill does not specify the use of delimiters or instructions to ignore embedded commands within the processed data.
- Capability inventory: Shell execution (CI steps), Git operations (worktrees, branching, cloning), GitHub CLI usage (PR viewing), and file system read/write access.
- Sanitization: Absent; there is no requirement for the agent to validate or sanitize the content of the specification or CI scripts before execution.
Audit Metadata