The Agent Skills Directory

[DATA_EXFILTRATION]: The skill is designed to read and parse .env files during documentation audits to ensure all environment variables are properly documented. This practice exposes sensitive credentials and secrets contained in these files to the agent's context.
[COMMAND_EXECUTION]: The instructions direct the agent to execute shell commands such as git log, git diff, and npm view to check for documentation freshness and dependency updates. It also uses filesystem commands like mkdir and mv for document management and archiving.
[INDIRECT_PROMPT_INJECTION]: The skill processes various untrusted files from the project directory, creating a surface for indirect prompt injection.
Ingestion points: Project configuration files (package.json, wrangler.jsonc), environment files (.env), and all markdown (.md) files in the project workspace.
Boundary markers: The skill does not define specific boundaries or instructions to ignore embedded commands within the content it reads from project files.
Capability inventory: The agent can write to and modify files, manage directories, and execute git and npm commands based on the analysis of these files.
Sanitization: There is no evidence of content sanitization or validation before the ingested data is used to influence the agent's output or documentation updates.
[EXTERNAL_DOWNLOADS]: The skill performs network requests to the npm registry via npm view to retrieve metadata about package versions. This targets a well-known service for legitimate documentation auditing purposes.

docs-workflow