The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user-supplied text as its primary function to identify and remove AI stylistic markers.
Ingestion points: The main text input provided by the user for the threads-humanizer command.
Boundary markers: The skill does not implement specific delimiters or instructions to isolate the user content during processing.
Capability inventory: The skill performs extensive regex transformation and interacts with external image generation and publishing services through the lib library.
Sanitization: The skill employs a comprehensive set of regex scrubbing rules defined in references/scrub-rules.md to identify and modify specific linguistic patterns.
[DYNAMIC_EXECUTION]: The skill provides Python-based logic templates in references/scrub-rules.md and sub-skills/illustration.md for text scoring, rhythm restoration, and image generation parameters. This logic is used by the agent to execute the humanization and illustration tasks.
[EXTERNAL_DOWNLOADS]: The skill utilizes external service integrations for image generation (Pixfaro), data retrieval (Apify), and social media publishing (Publora) as described in sub-skills/illustration.md and sub-skills/post-audit.md. These integrations are documented as part of the skill's intended functionality.

threads-humanizer