skill-creator
Pass
Audited by Gen Agent Trust Hub on Mar 5, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: Uses command execution to orchestrate internal scripts for benchmarking, grading, and packaging, as well as the Claude CLI for testing skill triggers. These are standard development operations.
- [SAFE]: Operates a local HTTP server restricted to localhost (127.0.0.1) for qualitative output review, with safe process management to clean up used ports.
- [SAFE]: Conducts automated description optimization by interacting with the Anthropic API using provided skill metadata, following standard LLM integration patterns.
- [SAFE]: No evidence of prompt injection, data exfiltration, or obfuscation was found in the instructions or the accompanying scripts.
Audit Metadata