The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill interacts with the Apify platform to retrieve LinkedIn post history via the apimaestro/linkedin-profile-posts actor. This interaction is essential for the skill's intended analytical purpose and includes explicit prompts to the user regarding the source, duration, and potential costs of the operation.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external social media profiles and project-specific files to perform scoring and generate rewrites. This configuration presents a risk where malicious instructions embedded in the ingested LinkedIn posts could attempt to influence the agent's behavior.
Ingestion points: LinkedIn post content scraped via Apify; about-me.md and voice.md project files.
Boundary markers: The instructions do not define clear delimiters or "ignore instructions" directives for the data processed during the analysis or rewriting phases.
Capability inventory: The skill possesses network access for tool invocation (Apify) and file system write capabilities for caching data ([username]-all-posts.json).
Sanitization: There are no explicit instructions to filter or sanitize external text before it is used to influence post rewriting or analysis output.

post-scorer