The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests and processes untrusted data from external X API endpoints (e.g., tweet text, user bookmarks, and timelines), which could contain malicious instructions.
Ingestion points: Data is retrieved via operations like get:/2/users/{id}/bookmarks and get:/2/users/{id}/timelines/reverse_chronological described in SKILL.md and references/usage-patterns.md.
Boundary markers: The instructions do not define boundary markers or explicit prompts to ignore instructions embedded within the retrieved data.
Capability inventory: The skill possesses significant capabilities including posting tweets (post:/2/tweets), managing authentication credentials, and establishing persistent streams via uxc subscribe.
Sanitization: There is no mention of sanitizing or filtering the content retrieved from the API before the agent processes it.
[COMMAND_EXECUTION]: The skill is built around executing shell commands using the uxc CLI and a dynamically created x-openapi-cli link.
Evidence: Numerous command examples in SKILL.md and references/usage-patterns.md demonstrate the use of uxc auth, uxc link, and uxc subscribe to perform operations.

x-openapi-skill