pidgin-share
Pass
Audited by Gen Agent Trust Hub on May 16, 2026
Risk Level: SAFEDATA_EXFILTRATIONPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [DATA_EXFILTRATION]: The skill is designed to read local files and upload them to the pidgin.sh API. While this is the intended purpose, it results in the transmission of user data to a remote service.\n- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection via its interactive 'response channel' feature, which instructs the agent to process external data.\n
- Ingestion points: JSON output from the
scripts/pidgin waitcommand.\n - Boundary markers: None. There are no instructions to differentiate between data and instructions in the external payload.\n
- Capability inventory: The agent has shell access and file interaction capabilities.\n
- Sanitization: No sanitization or validation of the remote payload is performed before the agent acts on it.\n- [COMMAND_EXECUTION]: The skill utilizes a Bash script to handle API interactions. The script uses
curlfor network requests and manages background tasks using temporary files in/tmp.
Audit Metadata