ca-django-tests
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: LOWCOMMAND_EXECUTIONDATA_EXFILTRATION
Full Analysis
- COMMAND_EXECUTION (LOW): The skill frequently executes shell commands using a local script
ca(e.g.,ca magictest,ca pytest,ca test). While these are intended for testing, they involve subprocess execution within the agent environment. - DATA_EXFILTRATION (LOW): The skill processes code from the
../ca/directory. While it does not explicitly exfiltrate data, its 'Indirect Prompt Injection' surface (Category 8) is present as it reads external code and test results. However, its capabilities are limited to local test execution and code modification, placing it in a lower risk tier. - INDIRECT PROMPT INJECTION (LOW): The skill ingests untrusted data (source code and test files from the filesystem). It uses this data to make decisions about test coverage and writing new tests. Since its primary side effect is writing Python tests and executing them via a local test runner, the risk is contained to the local development environment.
Audit Metadata