The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes a specialized CLI tool, postplus, to perform research tasks on X. It provides instructions for running search queries, retrieving profile information, and confirming spending quotes.\n
Evidence: Multiple shell command templates using the postplus executable are defined in SKILL.md and the references/ directory.\n- [INDIRECT_PROMPT_INJECTION]: The skill's primary function involves analyzing public data from X, which is an untrusted external source that may contain malicious instructions designed to influence the agent's behavior.\n
Ingestion points: External data enters the context via the x-posts, x-profiles, and x-user-search routes defined in SKILL.md and references/shared-contract.md.\n
Boundary markers: The skill includes logic to verify handles, check evidence quality, and use structured CLI output to avoid reliance on potentially malicious free-text fields (mentioned in SKILL.md under Public Command Boundary).\n
Capability inventory: The skill has the ability to execute shell commands (postplus), write results to files (--output), and approve spending (postplus quote confirm).\n
Sanitization: There are no explicit instructions or mechanisms defined to sanitize or filter the natural language content of posts and replies before they are analyzed by the agent.

x-research