The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands for project building, linting, and validation. Evidence: Instructions in the validation checklist guide the agent to use the Bash tool for running standard commands like npm run build, pytest, and specific internal validation scripts located in ~/.claude/skills/.
[INDIRECT_PROMPT_INJECTION]: The skill possesses a potential attack surface for indirect prompt injection as it processes data from external sources that could contain malicious instructions. Ingestion points: The agent reads from the contracts/ and shared/ directories, as well as README.md files, to understand API shapes and business logic. Capability inventory: The skill has the authority to write to the filesystem and execute shell commands via Bash. Sanitization: The instructions include specific security guidance to use createElement and textContent instead of innerHTML to prevent cross-site scripting (XSS). Boundary markers: The skill relies on the structure of machine-readable API contracts to constrain agent behavior.

frontend-agent