Add expected flow tests for remaining skills #36
Description
Summary
Expand expected flow test coverage from 1 to all 13 skills.
Current State
Only `tests/expected-flows/hitl-operators.yaml` exists, covering the analyzing-data skill for HITL operator queries.
Skills Needing Coverage
| Skill | Priority | Complexity |
|---|---|---|
| analyzing-data | ✅ Done | High |
| discover-data | High | Medium |
| debug-dag | High | Medium |
| dag-authoring | Medium | High |
| dag-testing | Medium | Medium |
| profile-table | Medium | Low |
| check-freshness | Low | Low |
| upstream-lineage | Low | Medium |
| downstream-lineage | Low | Medium |
| astro-project-setup | Low | Medium |
| astro-local-env | Low | Medium |
| init-warehouse | Low | Low |
| airflow-2-to-3-migration | Low | High |
Expected Flow Test Format
See `tests/expected-flows/hitl-operators.yaml` for the format:
- `name`: Test identifier
- `description`: What the test validates
- `queries`: Example user inputs
- `expected_flow`: Sequence of expected agent actions
- `success_criteria`: What defines a successful flow
- `failure_modes`: Known ways the flow can fail
Benefits
- Validates skill instructions work as intended
- Catches regressions when skills are updated
- Documents expected agent behavior
- Enables automated testing with mock responses
Files to Create
- `tests/expected-flows/discover-data.yaml`
- `tests/expected-flows/debug-dag.yaml`
- ... (one per skill)
Priority
Low - Important for quality but not blocking release
From PR #24 open-source readiness review