writing-great-skills
Pass
Audited by Gen Agent Trust Hub on Jul 19, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [NO_CODE]: The skill consists entirely of Markdown documentation (
SKILL.mdandGLOSSARY.md) and a configuration file (agents/openai.yaml). There are no scripts (Python, JavaScript, shell), binaries, or automation logic included. - [EXTERNAL_DOWNLOADS]: No external network requests, downloads, or remote package installations are defined in the files.
- [COMMAND_EXECUTION]: No shell commands, subprocesses, or system-level operations are present.
- [PROMPT_INJECTION]: The skill is a meta-guide on prompting. While it uses terms like 'instructions' and 'behavior,' it does not attempt to bypass safety filters or override the host agent's system prompt. It specifically teaches 'negation' and 'predictability' as safety and efficiency best practices.
Audit Metadata