vcr-http-recording
Pass
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: LOW
Full Analysis
- [SAFE] (SAFE): The skill provides templates and guidance for HTTP mocking in tests. No malicious code, prompt injection, or obfuscation was found.
- [DATA_EXFILTRATION] (SAFE): Although the skill captures network traffic, it includes robust filtering functions to redact sensitive headers, query parameters, and request/response body fields. It actively warns against committing real credentials to version control.
- [EXTERNAL_DOWNLOADS] (LOW): The skill references standard Python libraries including vcrpy, pytest, requests, and httpx. These are common industry-standard tools for the stated purpose.
- [INDIRECT_PROMPT_INJECTION] (LOW): While the skill processes external HTTP responses, it is used in a testing context and includes sanitization logic (redaction) to handle external data safely.
Audit Metadata