The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill provides a framework for analyzing and processing untrusted text from article drafts and external search results.
Ingestion points: Text content from user drafts and Search Engine Results Page (SERP) analysis data are ingested during the Architect, Carpenter, and Judge phases as described in SKILL.md and references/search-intent-analysis.md.
Boundary markers: The skill instructions and templates in SKILL.md lack explicit delimiters or instructions to treat ingested content as untrusted data.
Capability inventory: The skill is entirely instructional and does not utilize any code execution, file system access, or network tools.
Sanitization: There are no procedures defined for sanitizing or validating external input before processing.
[METADATA_POISONING]: The SKILL.md file identifies the author as https://github.com/dmitry, which conflicts with the provided author context of jeffallan. While not appearing to be malicious, this is an inconsistency in the skill's metadata.
[NO_CODE]: The skill contains only Markdown-based documentation and guidelines. No scripts, binaries, or automated tool configurations were detected in any of the provided files.

seo-writer