The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests and processes content from external sources which could potentially contain malicious instructions intended to manipulate the agent's behavior during report generation.
Ingestion points: The skill reads data from external GitHub repositories (integrations.github.repo), Figma files (integrations.figma.file_key), and Chromatic visual diffs.
Boundary markers: There are no explicit instructions or delimiters defined to prevent the agent from following natural language instructions found within the code or design specifications being analyzed.
Capability inventory: The skill is granted access to Write, Grep, Glob, and various Bash subcommands for file manipulation and analysis.
Sanitization: The instructions do not outline any sanitization or validation processes for the data retrieved from the integrated services.
[EXTERNAL_DOWNLOADS]: The skill is configured to interact with several well-known and official service providers to facilitate its analysis.
Figma: Fetches component specifications and properties using the Figma MCP.
GitHub: Searches repositories for code patterns, hardcoded values, and versioning data.
Chromatic: Pulls visual diff data to identify visual drift.
NPM: Uses npm view to verify the latest versions of design system packages.

drift-detection