The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to utilize shell commands for verifying the integrity and status of published content.
Uses curl -I to confirm that published URLs resolve with a 200 OK status.
Employs grep to perform a link count audit on the markdown body.
Uses diff to compare final publication payloads against the original user-supplied text.
[INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied content for publication, creating a potential vector for indirect prompt injection.
Ingestion points: The skill ingests "supplied copy" or "finished text" provided by the user in Supplied-copy mode as defined in SKILL.md.
Boundary markers: The skill explicitly labels this content as "approved copy" and "immutable," instructing the agent not to rewrite or replace the user's argument.
Capability inventory: The skill utilizes Bash, Write, Edit, and WebFetch tools to finalize and verify the publication.
Sanitization: The instructions recommend using "markdown formatting, escaping, [and] platform syntax" when handling the text, though the focus is on preserving the original message.
[EXTERNAL_DOWNLOADS]: The skill interacts with several external services and APIs for research and media retrieval.
Fetches image metadata and licensing information from the Wikimedia Commons API.
References SEO and keyword research tools including Ahrefs, Semrush, and DataForSEO for volume validation.
Mentions Google Search Console for retrieving actual search query data.

publish-article