twitter-reader
Pass
Audited by Gen Agent Trust Hub on Apr 7, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The Python script 'scripts/fetch_tweet.py' and the Bash script 'scripts/fetch_tweets.sh' use the 'curl' utility to fetch data from the Jina.ai Reader API. The Python script correctly uses a list of arguments with 'subprocess.run' to avoid shell injection vulnerabilities.
- [EXTERNAL_DOWNLOADS]: The skill connects to 'r.jina.ai' to retrieve tweet content in markdown format, which is the intended purpose of the tool.
- [PROMPT_INJECTION]: The skill processes untrusted data from external URLs, creating a surface for indirect prompt injection. 1. Ingestion points: Content is fetched from user-provided URLs in 'scripts/fetch_tweet.py' and 'scripts/fetch_tweets.sh'. 2. Boundary markers: The fetched content is returned as markdown without specific delimiters. 3. Capability inventory: The skill uses 'curl' via subprocess to retrieve external data. 4. Sanitization: Input URLs are validated to start with 'https://x.com/' or 'https://twitter.com/'.
Audit Metadata