permacc
Pass
Audited by Gen Agent Trust Hub on Sep 7, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes a local Python script (
scripts/permacc.py) via theuv runtool to perform its primary functions. - [EXTERNAL_DOWNLOADS]: The Python script specifies
requestsas a dependency, which is downloaded and installed byuvat runtime.requestsis a standard, well-known library for HTTP operations. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data in the form of URLs and JSON inventory files, creating a potential surface for indirect prompt injection.
- Ingestion points: Data is ingested through positional command-line arguments and file inputs in
scripts/permacc.py. - Boundary markers: No specific markers or delimiters are used to isolate untrusted content or instruct the agent to ignore instructions embedded in the data.
- Capability inventory: The script performs network requests (GET, POST, PATCH, DELETE) to the Perma.cc API and performs local file read/write operations.
- Sanitization: The skill does not perform input validation or sanitization on the provided URLs, relying on the target API's own handling.
Audit Metadata