The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests data from external sources to influence code generation and validation steps, creating a potential surface for instructions embedded in documentation to be executed by the agent.
Ingestion points: The skill retrieves information using MCP tools get_usage_guide, get_component_api, and get_component_examples, as well as reading the local file libs/mcp-server/src/data/usage-guides.ts.
Boundary markers: No specific delimiters or warnings are used to instruct the agent to ignore potentially malicious instructions embedded within the ingested usage guides or examples.
Capability inventory: The agent has access to Write and Edit tools to create/modify source code, and the Bash(nx *) tool to execute build commands.
Sanitization: The skill does not implement explicit sanitization or validation logic for the content retrieved from the MCP server or documentation files before using it to generate component templates or logic.
[COMMAND_EXECUTION]: The skill instructions include the execution of shell commands to validate the generated components.
Evidence: Phase 6 explicitly directs the agent to run nx run <project>:build or ng build to verify compilation. While these are restricted to the nx and ng CLIs in a development context, they involve executing complex build pipelines based on generated code.

scaffold