The Agent Skills Directory

[SAFE]: The skill is entirely instructional and provides a workflow for an AI agent to ground its architectural advice in source files. It does not contain any executable scripts, binary files, or automated commands.- [PROMPT_INJECTION]: The instructions focus on improving recommendation quality and confidence labeling. There are no attempts to bypass safety guardrails, override system prompts, or elicit restricted behaviors.- [DATA_EXFILTRATION]: No network operations, hardcoded credentials, or sensitive file path access (e.g., SSH keys, environment secrets) were detected. The skill's focus on reading codebase files is restricted to architectural analysis within the agent's established context.- [REMOTE_CODE_EXECUTION]: There are no patterns involving remote script downloads, package installations, or dynamic code execution. The reference to 'agentic-evals' in the YAML frontmatter represents a standard internal skill composition pattern rather than an untrusted external dependency.

grounded