Loading...
Loading...
Write and improve agent skills (SKILL.md files). Use when creating new skills, refactoring existing ones, debugging why an agent ignores instructions, or improving compliance. Covers prompt structure, TDD for skills, description optimization, and rationalization bulletproofing.
npx skill4agent add millionco/expect skill-writing| TDD Step | Skill Equivalent |
|---|---|
| RED | Run scenario without skill. Document what the agent does wrong. |
| GREEN | Write minimal skill addressing those specific failures. Verify agent now complies. |
| REFACTOR | Find new rationalizations the agent invents. Close loopholes. Re-test. |
accessibility_auditRUN_COMPLETEDdescriptiondescription: A skill for TDD - write test first, watch it fail, write codedescription: Use when tests have race conditions, timing dependencies, or pass/fail inconsistently. Covers flaky tests, hanging tests, zombie processes..agents/skills/<skill-name>/
SKILL.md # Required. Under 200 lines.
references/ # Optional. Heavy reference docs (100+ lines).---
name: kebab-case-name
description: Use when [triggering conditions]. Covers [symptoms, tools, keywords].
---creating-skillsskill-creationcondition-based-waitingasync-helpers# Browser Test Executor
You are executing adversarial browser tests against code changes.
REQUIRED for every test run:
1. Call accessibility_audit before completing
2. Call performance_metrics before completing
3. Call close to flush the sessionBefore marking the task complete, you MUST:
1. Run the linter
2. Run the type checker
3. Run the test suite
Do not skip any step. A skipped step is a failed task.call accessibility_auditrun pnpm typecheck**Bad:** `test the login page`
**Good:** `Submit login with empty fields, invalid email, wrong password, valid credentials. Verify errors and redirect.`| Agent Says | Skill Responds |
|---|---|
| "This change is too small to test" | Every change gets tested. Size is not an exemption. |
| "I already verified manually" | Manual verification is not a substitute. Run the tool. |
| "The user didn't ask for this" | The skill requires it. User silence is not opt-out. |
| Pattern | Why It Fails |
|---|---|
| "Consider running X when appropriate" | Conditional + vague = ignored |
| Giant skill files (>200 lines) | Exceeds instruction budget, compliance drops |
| Duplicating CLAUDE.md rules | Burns budget twice for the same instruction |
| Listing every edge case | Show one example, not twenty — the agent infers the rest |
| "Be thorough" / "Be careful" | Vibes, not instructions. Say what "thorough" means concretely |
| Multi-paragraph justifications | The agent needs the rule, not the backstory. One sentence of "why" max |
| Critical rules in the middle | Middle-of-prompt blindspot. Move to top or bottom |
| Writing a skill without testing first | You don't know what failures to address. Delete it, start with RED. |
| Workflow-summarizing descriptions | Agent follows the summary instead of loading the full skill |