categorizing-bsky-accounts

Warn

Audited by Gen Agent Trust Hub on Apr 26, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The script scripts/bluesky_analyzer.py performs dynamic execution by generating a Python script as a string and running it through a subprocess call to a virtual environment interpreter.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection through account data. 1. Ingestion points: Data is fetched from the Bluesky API in scripts/bluesky_analyzer.py. 2. Boundary markers: There are no markers or safety instructions delimiting the untrusted account content in the output. 3. Capability inventory: The skill utilizes subprocess execution and network requests. 4. Sanitization: No filtering or validation is performed on user-controlled bios or posts before they are presented to the agent.
  • [EXTERNAL_DOWNLOADS]: Fetches data from https://public.api.bsky.app, which is the official public API for the Bluesky social network, for legitimate account analysis purposes.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Apr 26, 2026, 06:39 AM