Enforce Test-Driven Development cycles for creating and modifying AI skills.
npx skills add https://github.com/ZJUxjy/opencode_lite --skill writing-skills-zjuxjy
Please help me install this Agent Skill. Skill: writing-skills Source: https://github.com/ZJUxjy/opencode_lite/tree/main/skills/writing-skills Command: npx skills add https://github.com/ZJUxjy/opencode_lite --skill writing-skills-zjuxjy
This Skill ensures that new and updated skills are rigorously tested and documented, preventing common errors and rationalizations that lead to ineffective or bypassed skills.
Follow the RED-GREEN-REFACTOR process outlined in this skill's documentation to test your new skill.
High-intent search queries and answers about installing and using this skill.
To test agent compliance for skill documentation, apply Test-Driven Development principles using RED-GREEN-REFACTOR cycles and pressure scenarios with subagents to verify rules are followed under stress.
The best way to prevent agents from bypassing skill rules is implementing explicit rationalization countermeasures during skill creation, closing loopholes agents use to skip established protocols.
Apply TDD to skill development by adapting the RED-GREEN-REFACTOR cycle: write failing tests for agent compliance, create the skill to pass tests, and refactor until agents comply under pressure.
AI skills fail under pressure scenarios when documentation lacks rigorous testing against combined pressures like time constraints, sunk cost, and authority, allowing agents to rationalize bypassing rules.
Yes, you need subagents to simulate real-world stress during pressure scenario testing, verifying that your skill enforces coding standards and maintains agent compliance under combined pressures.
Use test-driven skill development before deploying any new skill that enforces critical coding standards, ensuring the skill is rigorously tested and documented against agent rationalization and bypass attempts.
ucirelloApply Test-Driven Development cycles to create and refine Skill documentation.
kazukitcyManage agent skill lifecycles through creation, improvement, and pruning workflows with regression testing.
obraApply RED-GREEN-REFACTOR cycles to create and test AI skills.
0xUrsanomicsAudit agent skill artifacts for malicious threat chains and structural vulnerabilities.
