skill-authoring
Pass
Audited by Gen Agent Trust Hub on Aug 4, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to run a local repository script
./scripts/lint-skills.shto validate skill directory structure and file permissions. - [COMMAND_EXECUTION]: Includes instructions for using the
tesslCLI tool, including commands for skill review (tessl skill review) and automated optimization (tessl skill review --optimize). - [EXTERNAL_DOWNLOADS]: Interacts with an external scoring service via the
tesslCLI, which requires user authentication (tessl login) to access advanced features like LLM-based optimization. - [DATA_EXFILTRATION]: The auditing workflow involves sending the content of skill files (SKILL.md and references) to the Tessl service for evaluation. This is a documented and functional requirement for the skill's stated goal of providing quality scores.
- [PROMPT_INJECTION]: The skill guidelines teach 'pushy' description patterns designed to improve agent triggering performance. While these use imperative language to influence agent behavior, they are presented as quality best practices rather than safety bypasses.
- [SAFE]: References external documentation and resources from trusted sources, including Anthropic's official platform and public GitHub repositories.
Audit Metadata