The Agent Skills Directory

[SAFE]: No malicious patterns were detected. The skill implements several robust safety controls:
Input Sanitization: The instructions explicitly direct the agent to treat all user-provided tag names as untrusted input, mitigating potential shell injection risks.
Credential Security: It prevents the exfiltration of sensitive tokens by instructing users to manage auth via environment variables or local CLI commands rather than sharing them in the chat context.
Operation Safety: The skill enforces a two-step verification process for write operations (create, rename, move, merge), requiring a --dry-run preview and explicit user approval before execution.
[COMMAND_EXECUTION]: The skill uses the Bash tool to execute the dino CLI. This behavior is consistent with the skill's documented purpose and is constrained by the allowed-tools configuration.
[PROMPT_INJECTION]: The skill processes user-supplied tag names as input for CLI commands, creating an indirect injection surface. However, this is mitigated by mandatory boundary instructions and user-in-the-loop confirmation for all state-changing actions.
Ingestion points: User-provided tag names and paths in SKILL.md (e.g., dino tag add [name]).
Boundary markers: Presence of explicit "Safety & Boundaries" section requiring untrusted input handling.
Capability inventory: Shell command execution via Bash tool.
Sanitization: Reliance on agent-side input validation and CLI-level dry-runs.

dino-manage-tags