The Agent Skills Directory

[PROMPT_INJECTION]: The skill interacts with external content from X/Twitter (tweets, bios, and messages) which presents a surface for indirect prompt injection.
Ingestion points: Multiple endpoints described in references/x-api.md (e.g., /x/tweets/search, /x/users/{id}).
Boundary markers: The skill mandates the use of <XQUIK_UNTRUSTED_X_CONTENT> delimiters as specified in SKILL.md and references/security.md.
Capability inventory: Limited to Xquik API requests; no shell execution or local file system access is enabled.
Sanitization: Clear instructions are provided to treat all retrieved content as data only and to ignore any instructions found within the data.
[EXTERNAL_DOWNLOADS]: The skill communicates with official Xquik domains (xquik.com and docs.xquik.com) for API functionality and documentation retrieval. These references to a well-known service are documented neutrally and align with the skill's primary purpose.
[CREDENTIALS_UNSAFE]: The skill correctly manages authentication by using environment variables for the XQUIK_API_KEY. It contains explicit prohibitions against requesting or handling sensitive X account credentials like passwords, 2FA codes, or session cookies.
[COMMAND_EXECUTION]: No unauthorized or dangerous command execution patterns were detected. The skill is designed for HTTPS API interactions and explicitly states in its security guardrails that it does not run shell commands.

x-twitter-scraper