The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill identifies and reads content from all markdown and HTML files within the project's blog directories, creating a vector for indirect instructions.
Ingestion points: The audit process reads files with *.md, *.mdx, and *.html extensions from directories like content/, posts/, and blog/ as specified in SKILL.md.
Capability inventory: The skill is granted access to Bash, Write, and Task tools, which could be misused if the agent's logic is subverted by malicious content.
Boundary markers: The instructions do not specify the use of delimiters or clear separation between system instructions and the ingested blog content when spawning subagents for analysis.
Sanitization: There is no evidence of content filtering or sanitization to prevent the agent from interpreting instructions embedded within the blog posts being audited.

blog-audit