Writing Skills

Use when creating new skills, editing existing skills, or verifying skills work before deployment

Reasonable to install if you want help authoring skills. Review any generated or edited skill files before keeping them because they can steer future agents, keep pressure-testing prompts confined to controlled testing, and only run render-graphs.js on skill directories and Graphviz installations you trust.

Type: OpenClaw Skill Name: writing-skills Version: 0.1.0 The skill bundle is designed for developing and testing OpenClaw skills, focusing on a Test-Driven Development (TDD) approach for documentation. The `SKILL.md` and related markdown files (`persuasion-principles.md`, `testing-skills-with-subagents.md`) contain extensive meta-instructions for the AI agent on how to author, structure, and bulletproof skills against agent 'rationalization' (deviation from instructions). While these instructions aim to strongly control the agent's behavior, they are strictly within the benign context of improving skill development and compliance, not for harmful objectives like data exfiltration or unauthorized access. The `render-graphs.js` script uses `execSync` to run the `dot` (Graphviz) command-line tool, which is a high-risk function, but its usage is confined to rendering diagrams from `SKILL.md` content, a legitimate and stated purpose. No evidence of intentional harmful behavior, data exfiltration, persistence, or unauthorized command execution was found.