The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data in the form of user-provided PageSpeed Insights reports or URLs.
Ingestion points: The workflow begins in SKILL.md and references/detailed-guide.md when a user pastes a performance report or provides a URL for scanning.
Boundary markers: There are no explicit delimiters or instructions for the agent to ignore instructions that might be embedded within the report content.
Capability inventory: The skill possesses significant capabilities across its scripts, including modifying configuration files (netlify.toml, vercel.json, vite.config.ts), HTML/CSS files, and executing shell commands for image optimization (cwebp, sips, python3).
Sanitization: The instructions do not specify any validation or sanitization of the input data before it is processed by the agent.
[DYNAMIC_EXECUTION]: The skill utilizes dynamic script execution for asset optimization tasks.
Evidence: In references/detailed-guide.md, the skill provides python3 -c command templates to perform image resizing and icon optimization using the Pillow library.
Context: These are low-risk, template-based executions intended for local file processing and do not incorporate direct untrusted input into the command string.

pagespeed-enhancer