crank-test-prune
Pass
Audited by Gen Agent Trust Hub on Oct 22, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes test files to determine verdicts, creating a surface for adversarial content to influence the agent. Ingestion points: Test files identified in the repository inventory (SKILL.md, Step 1). Boundary markers: No specific markers or instructions to ignore embedded content are defined. Capability inventory: File deletion and modification (SKILL.md, Step 3) and execution of the suite's run command (SKILL.md, Step 4). Sanitization: No validation or sanitization of test file content is performed.
- [COMMAND_EXECUTION]: The skill runs the project's test suite command in a temporary worktree. This is a capability to execute shell commands which could be exploited if the test command itself is malicious.
Audit Metadata