The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze external artifacts such as test code, application source files, and test result data (e.g., playwright-results.json). While this creates a surface for indirect prompt injection where malicious content in these files could influence agent behavior, the risk is inherent to the primary function of a testing automation skill.
Ingestion points: The skill reads source code, test suites, and structured test results from the repository.
Boundary markers: No explicit delimiters are specified for separating untrusted data from instructions, though it encourages following project-specific conventions in CLAUDE.md.
Capability inventory: The agent has access to shell execution (Bash), file system modification (Write, Edit), and task management tools.
Sanitization: The instructions do not describe sanitization or filtering of external data before processing by the agent.
[EXTERNAL_DOWNLOADS]: The skill uses standard package managers and tools (npm, npx, pip) to install and run testing frameworks like Playwright and Vitest. These operations target well-known, trusted registries and services.
[COMMAND_EXECUTION]: The skill utilizes shell commands to execute test suites and quality gates. All commands are standard development tools and are used within the context of the skill's stated purpose.

testing