The Agent Skills Directory

[Prompt Injection] (SAFE): No patterns detected that attempt to override system prompts, bypass safety filters, or ignore previous instructions. The 'Use when' clauses are standard functional triggers.
[Data Exposure & Exfiltration] (SAFE): No hardcoded credentials, API keys, or sensitive file paths were found. The skill does not perform any network operations.
[Unverifiable Dependencies] (SAFE): The skill does not install any packages from npm, pip, or other registries, nor does it download remote scripts.
[Indirect Prompt Injection] (LOW): The skill instructs the agent to read local reference files (index.md, frontmatter.md, etc.) to inform the creation of new agents. While this constitutes an ingestion surface, the risk is negligible as it pertains to local project documentation and the skill lacks automated execution logic.
[Privilege Escalation] (SAFE): No commands like sudo or chmod are present. Tool declarations (Bash, Edit) are provided as template examples for agents being designed and do not represent active escalation by the skill itself.

agent-builder