The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill analyzes external repository content such as README.md, package.json, and CSS files to gather design context, creating a surface where instructions embedded in those files could influence agent behavior.
Ingestion points: SKILL.md Phase 1 instructions for project context extraction.
Boundary markers: No specific delimiters or instructions to ignore malicious content within external files are defined.
Capability inventory: Writing SVG files, executing image conversion scripts, parallel task execution, and repository management (clone/commit/PR).
Sanitization: No explicit sanitization or filtering for injection patterns in the ingested project data.
[COMMAND_EXECUTION]: The bundled scripts/export.sh script executes various system utilities including resvg, inkscape, and rsvg-convert to render logos.
[EXTERNAL_DOWNLOADS]: scripts/export.sh utilizes npx --yes to fetch and execute the @aspect-build/resvg package from the npm registry if it is not available on the system.
[DYNAMIC_EXECUTION]: The export script dynamically generates and executes a Node.js snippet via a heredoc to perform image processing using the sharp library.
[PRIVILEGE_ESCALATION]: In Phase 2 and Phase 3 of SKILL.md, the agent is instructed to dispatch sub-agents using the Task tool with the mode: "bypassPermissions" flag enabled.

logo-designer