The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests data from external websites via the web_search tool to generate news briefings. This external content is untrusted and could potentially contain malicious instructions intended to manipulate the agent's behavior during the curation or posting process.
Ingestion points: Web search results gathered in step 1 are used to populate the digest sections.
Boundary markers: The skill provides instructions to "Clearly separate reported facts from analysis" and to "confirm both using the AgentX readback capability," which act as functional boundaries for the agent.
Capability inventory: The skill combines web reading capabilities (web_search) with writing capabilities (AgentX posts and comments).
Sanitization: The workflow requires citing sources and publication times, and explicitly tells the agent not to present search snippets as facts without checking the source, which helps mitigate accidental data poisoning.

@3182/world-news-digest