shinka-convert
Pass
Audited by Gen Agent Trust Hub on Mar 12, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill implements a smoke-testing phase where it executes the newly generated
evaluate.pyscript using the local Python interpreter to verify the task configuration. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted files from the current working directory to create code snapshots and generate evaluation logic, creating a surface where source code content influences agent-generated scripts.
- Ingestion points: Inspects and snapshots the current working directory to identify entrypoints and dependencies (documented in SKILL.md workflow).
- Boundary markers: No explicit markers are used to delimit untrusted source code during the generation of the
evaluate.pycontract. - Capability inventory: The skill performs file system writes to create the
./shinka_task/directory and executes subprocesses during the smoke test (documented in SKILL.md step 8). - Sanitization: There is no evidence of content validation or escaping of the ingested source code before it is rewritten into the Shinka contract.
Audit Metadata