Agent skill
aico-pm-acceptance-criteria
Define clear, testable acceptance criteria using Given/When/Then (Gherkin) format that can be directly used for testing. Use this skill when: - User asks for "acceptance criteria", "AC", "test criteria", "done criteria" - User asks "how do we know it's done?", "what counts as complete?" - User story needs testable success conditions - QA needs test scenarios or test cases - Running /pm.plan and need to add AC to stories - Reviewing story and AC is missing or vague Format: Always use Given [context], When [action], Then [expected result] structure.
Install this agent skill to your Project
npx add-skill https://github.com/majiayu000/claude-skill-registry/tree/main/skills/other/aico-pm-acceptance-criteria
SKILL.md
Acceptance Criteria
⚠️ CRITICAL RULES - READ FIRST
- READ STORY FIRST: Always read the story file from
docs/reference/pm/stories/before writing criteria - USE GIVEN/WHEN/THEN: All criteria must follow this format
- UPDATE STORY FILE: Add criteria to existing story file, don't create new files
Language Configuration
Before generating any content, check aico.json in project root for language field to determine the output language. If not set, default to English.
Process
- Review story/feature: Understand what needs to be done
- Identify success scenarios: Happy paths first
- Write Given/When/Then: For each scenario
- Add edge cases: Boundary conditions and errors
- Verify testability: Each criterion must be independently verifiable
Acceptance Criteria Template
### Acceptance Criteria for [Story/Feature]
#### Scenario 1: [Scenario Name]
- **Given** [precondition/context]
- **When** [action/trigger]
- **Then** [expected outcome]
#### Scenario 2: [Scenario Name]
- **Given** [precondition/context]
- **When** [action/trigger]
- **Then** [expected outcome]
#### Edge Cases
- **Given** [edge case condition]
- **When** [action]
- **Then** [expected behavior]
Criteria Categories
| Category | Examples |
|---|---|
| Functional | Feature works as specified |
| Validation | Input validation rules |
| Error handling | Error messages and recovery |
| Performance | Response time expectations |
| Accessibility | A11y requirements |
Quality Checklist
- Each criterion is independently testable
- No ambiguous language ("should", "might", "could")
- Edge cases are covered
- Error states are defined
- Performance expectations stated (if relevant)
Key Rules
- ALWAYS use Given/When/Then structure
- MUST make each criterion independently testable
- ALWAYS include edge cases and error scenarios
- NEVER use vague language - be specific
Common Mistakes
- ❌ Vague criteria ("works correctly") → ✅ Specific conditions
- ❌ Missing edge cases → ✅ Include boundary conditions
- ❌ Implementation details → ✅ Focus on observable behavior
Recommended Agent Skills
Expand your agent's capabilities with these related and highly-rated skills.
agent-ops-spec
Manage specification documents in .agent/specs/. Use when user provides requirements, acceptance criteria, or feature descriptions that need to be tracked and validated against implementation.
agent-ops-state
Maintain .agent state files. Use at session start, after meaningful steps, and before concluding: read/update constitution/memory/focus/issues/baseline consistently.
agent-ops-spec
Manage specification documents in .agent/specs/. Use when user provides requirements, acceptance criteria, or feature descriptions that need to be tracked and validated against implementation.
agent-ops-testing
Test strategy, execution, and coverage analysis. Use when designing tests, running test suites, or analyzing test results beyond baseline checks.
agent-ops-testing
Test strategy, execution, and coverage analysis. Use when designing tests, running test suites, or analyzing test results beyond baseline checks.
agent-ops-state
Maintain .agent state files. Use at session start, after meaningful steps, and before concluding: read/update constitution/memory/focus/issues/baseline consistently.
Didn't find tool you were looking for?