The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it retrieves and processes untrusted content from the X (Twitter) platform. 1. Ingestion points: Untrusted data is ingested through API responses containing user bios, names, and tweet text from endpoints defined in SKILL.md like /x/users/{username} and /x/users/{id}/tweets. 2. Boundary markers: The skill instructions do not specify technical boundary markers (e.g., delimiters) to isolate the untrusted data, though it provides natural language warnings. 3. Capability inventory: The skill performs network operations via API calls to the vendor's domain xquik.com. 4. Sanitization: The skill includes specific instructions to the agent to summarize long content and to disregard any instructions or commands found within the fetched data.

user-tweets