The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes local source files and existing test files to generate instructions for the eval-test-creator agent. This creates an attack surface where malicious content inside the project files could influence the agent's behavior.
Ingestion points: Reads source code from src/umbraco-api/tools/{collection}/index.ts and existing scenarios from tests/evals/*.test.ts.
Boundary markers: No explicit delimiters or instructions are provided to the agent to treat the ingested code as untrusted data.
Capability inventory: The agent can create new test files and execute shell commands via npm scripts.
Sanitization: No sanitization or validation of the file content is performed before it is used to influence test creation.
[COMMAND_EXECUTION]: The skill instructs the agent to run npm run build and npm run test:evals. These are standard development operations but execute code based on local configurations.

add-eval