common-skill-creator
Pass
Audited by Gen Agent Trust Hub on Oct 10, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is entirely instructional, providing a standardized workflow for developing other agent skills. It does not contain executable code, remote script downloads, or obfuscated content.
- [PROMPT_INJECTION]: No malicious prompt injection patterns were detected. On the contrary, the skill includes a 'Pressure Hardening' protocol and behavior guardrails designed to identify and mitigate agent 'laziness' and potential prompt shortcuts in other skills.
- [COMMAND_EXECUTION]: The skill refers to standard development commands such as
cp -rfor file versioning (snapshots) andpnpmfor running evaluation suites. These are local operations intended for use in a controlled development workspace. - [INDIRECT_PROMPT_INJECTION]: While the skill involves processing user-provided descriptions to generate new skill files, it promotes secure practices by encouraging the use of clear delimiters, anti-patterns, and rigorous assertion-based testing to verify agent behavior.
Audit Metadata