fetch-tweets
Pass
Audited by Gen Agent Trust Hub on Aug 25, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill contains a vulnerability surface for indirect prompt injection by interpolating unvalidated user input into model prompts.\n
- Ingestion points: User input from the
${var}parameter is parsed and stored in theARGvariable (SKILL.md).\n - Boundary markers: The skill lacks explicit delimiters or negative constraints to prevent the model from following instructions embedded within the user's search keywords.\n
- Capability inventory: The agent can execute network requests via
curl, write to the local file system, and send notifications to the user.\n - Sanitization: While numeric validation is performed for list IDs and regex for account handles, the
keywordbranch passes search terms to the X.AI API verbatim.\n- [COMMAND_EXECUTION]: The skill performs shell command execution usingbashto orchestrate data fetching and processing.\n - Evidence: Use of
curl,jq, and local utilities like./secretcurland./notifyto interact with external APIs and the user (SKILL.md).
Audit Metadata