The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted external content from LinkedIn comments, which presents a surface for instructions embedded in those comments to influence agent behavior.
Ingestion points: Data is fetched via lib.ApifyClient.fetch_post_comments and lib.url_parser.parse_linkedin_url as documented in SKILL.md.
Boundary markers: The skill contains a dedicated 'Untrusted content' section in SKILL.md that explicitly instructs the agent to treat fetched text as data and never as instructions.
Capability inventory: The skill has the ability to post reactions and replies to LinkedIn via lib.publish.
Sanitization: The skill employs a stylistic 'humanizer' pass (linkedin-humanizer), but relies primarily on instructional guardrails to handle adversarial input.
[COMMAND_EXECUTION]: The skill invokes linkedin-humanizer to refine the voice of the generated drafts. This is a standard component of the author's social media toolset.
[EXTERNAL_DOWNLOADS]: The skill retrieves data from LinkedIn using the Apify platform, an established external service.

linkedin-reply-handler