The Agent Skills Directory

[Behavioral Templates] (SAFE): The skill is a collection of system prompt extensions that define how an agent should format its responses and prioritize tasks in different modes (e.g., BRAINSTORM, IMPLEMENT). It contains no logic that attempts to bypass safety filters or override core agent constraints.
[Tool Access] (SAFE): The YAML frontmatter restricts allowed tools to 'Read', 'Glob', and 'Grep'. This read-only profile prevents the skill from modifying the system or exfiltrating data via the network.
[Indirect Prompt Injection Surface] (SAFE): The skill is designed to analyze external data such as code and logs in DEBUG and REVIEW modes. While this constitutes an ingestion surface for indirect prompt injection, the risk is negligible due to the absence of high-impact tools (e.g., shell execution, network requests, or file writing). Evidence Chain: 1. Ingestion points: External files and logs are processed via 'Read' and 'Grep' tools. 2. Boundary markers: Absent; the skill does not explicitly instruct the agent to distinguish between its instructions and content within processed files. 3. Capability inventory: Limited to read-only filesystem inspection. 4. Sanitization: No sanitization logic is present for data ingested from the filesystem.

behavioral-modes