@tgoodington/intuition 11.4.0 → 11.5.1

package/docs/archive/QUICK_TEST_CHECKLIST.md

@@ -1,467 +0,0 @@
-# Quick Testing Checklist - Waldo v3 Validation
-
-**Use this checklist to verify critical functionality during testing.**
-
-**Version:** 3.0.0 | **Date:** February 5, 2026
-
----
-
-## Pre-Testing Validation
-
-- [ ] Version in package.json is 3.0.0
-- [ ] All skill files present (start, initialize, discovery, handoff, plan, execute)
-- [ ] WALDO_V3_COMPLETE_DOCUMENTATION.md exists and is readable
-- [ ] .claude/USER_PROFILE.json template exists
-- [ ] Skills can be activated (npm install -g intuition)
-
----
-
-## Test Session 1: Waldo v3 Core Features
-
-### Setup
-- [ ] Create test project directory
-- [ ] Run `/intuition-initialize` to set up project memory
-- [ ] Verify docs/project_notes/ created with all templates
-
-### Dialogue Mode Selection
-- [ ] Run `/intuition-discovery`
-- [ ] Waldo greets with mode selection question
-- [ ] Guided mode option available
-- [ ] Open-Ended mode option available
-- [ ] Can select mode without error
-- [ ] Mode stored in .project-memory-state.json > discovery.dialogue_mode
-
-### Research Agent Launch
-- [ ] User describes context
-- [ ] Waldo identifies domain from context
-- [ ] Waldo launches 3 research tasks (check Task tool calls)
-- [ ] All 3 research tasks run in parallel (not sequential)
-- [ ] Research completes while dialogue continues
-- [ ] Research findings are available in state
-
-### Guided Mode Dialogue
-- [ ] Questions use AskUserQuestion format
-- [ ] 2-4 options presented per question
-- [ ] "Other" option always available
-- [ ] Never more than 2 questions per turn
-- [ ] Questions informed by research findings
-- [ ] "Yes, and..." building evident in dialogue
-- [ ] Gentle steering (if applicable) is non-prescriptive
-
-### Open-Ended Mode Dialogue
-- [ ] Questions are conversational (no options)
-- [ ] Natural dialogue flow
-- [ ] Never more than 2 questions per turn
-- [ ] Questions informed by research findings
-- [ ] Covers same GAPP dimensions as Guided mode
-- [ ] Same depth and rigor as Guided mode
-
-### Discovery Completion
-- [ ] All GAPP dimensions explored (Problem, Goals, Context, Motivation)
-- [ ] Assumptions documented with confidence levels
-- [ ] User profile information discovered naturally
-- [ ] discovery_brief.md created with complete structure
-- [ ] discovery_output.json created with structured data
-- [ ] user_profile_learnings populated in discovery_output.json
-
----
-
-## Test Session 2: Handoff Orchestration
-
-### User Profile Extraction & Merging
-- [ ] Handoff reads discovery_output.json
-- [ ] Extracts user_profile_learnings
-- [ ] .claude/USER_PROFILE.json created or updated
-- [ ] Null fields populated from discovery learnings
-- [ ] Existing fields NOT overwritten (unless high confidence)
-- [ ] last_updated timestamp updated
-- [ ] projects_contributed_to includes current project
-- [ ] confidence_scores updated for each property
-
-### Memory Updates
-- [ ] key_facts.md updated with discovered facts
-- [ ] Facts include dates and sources
-- [ ] decisions.md updated with new ADRs (if applicable)
-- [ ] ADRs include Context/Decision/Consequences
-- [ ] issues.md updated with work logged
-- [ ] All updates use proper formatting
-
-### Brief Generation
-- [ ] planning_brief.md created (if discovery→planning)
-- [ ] planning_brief includes all required sections
-- [ ] Brief is focused and actionable for Magellan
-- [ ] execution_brief.md created (if planning→execution)
-- [ ] execution_brief includes all required sections
-- [ ] Brief is focused and actionable for Faraday
-
-### State Transition
-- [ ] workflow.status updated correctly
-- [ ] Timestamps for phase starts/completes updated
-- [ ] discovery.completed = true (after discovery handoff)
-- [ ] planning.started = true (after discovery handoff)
-- [ ] All state changes consistent
-
----
-
-## Test Session 3: Dual Dialogue Mode Equivalence
-
-### Setup
-- [ ] Run /intuition-discovery in Guided mode (Project A)
-- [ ] Run /intuition-discovery in Open-Ended mode (Project B)
-- [ ] Use same context/topic for both
-
-### Comparison
-- [ ] discovery_brief.md content is substantially same
-- [ ] GAPP coverage is equivalent
-- [ ] Assumptions are equivalent
-- [ ] Tone differs (options vs. conversational)
-- [ ] Depth and rigor are equivalent
-- [ ] Both route to /intuition-handoff
-
----
-
-## Test Session 4: User Profile Persistence
-
-### Setup
-- [ ] Complete discovery in Project A
-- [ ] Handoff merges profile
-- [ ] Check .claude/USER_PROFILE.json has data
-
-### Verify Persistence
-- [ ] Start Project B
-- [ ] .claude/USER_PROFILE.json already populated
-- [ ] Magellan reads existing profile (if testing planning)
-- [ ] Faraday reads existing profile (if testing execution)
-- [ ] Run discovery in Project B
-- [ ] New learnings merged into profile
-- [ ] Profile completeness increases
-- [ ] Projects list includes both projects
-
----
-
-## Test Session 5: Five-Skill Coordination
-
-### Full Flow (Discovery → Plan → Execution)
-- [ ] /intuition-start shows discovery not started
-- [ ] /intuition-discovery creates discovery outputs
-- [ ] /intuition-start shows discovery complete
-- [ ] /intuition-handoff processes discovery
-- [ ] /intuition-start shows planning ready
-- [ ] /intuition-plan reads planning_brief.md
-- [ ] /intuition-plan creates plan.md
-- [ ] /intuition-plan routes to user approval
-- [ ] /intuition-handoff processes planning
-- [ ] /intuition-start shows execution ready
-- [ ] /intuition-execute reads execution_brief.md
-- [ ] /intuition-execute completes workflow
-
-### State Consistency
-- [ ] .project-memory-state.json updated at each phase
-- [ ] No data loss between phases
-- [ ] File paths consistent
-- [ ] Timestamps accurate and ordered
-
----
-
-## Test Session 6: Resume Capability
-
-### Interrupt & Resume Discovery
-- [ ] Start /intuition-discovery
-- [ ] Partial dialogue (2-3 turns)
-- [ ] Interrupt (don't complete)
-- [ ] Check state tracking
-- [ ] Run /intuition-discovery again
-- [ ] Waldo resumes from last turn
-- [ ] Can complete discovery
-- [ ] Quality score shows same coverage
-
-### Interrupt & Resume Handoff
-- [ ] Run /intuition-handoff (partial execution simulated)
-- [ ] Interrupt mid-way through updates
-- [ ] Run /intuition-handoff again
-- [ ] Resumes and completes
-- [ ] No duplicate updates in memory
-
----
-
-## Test Session 7: File Management & State
-
-### File Creation & Integrity
-- [ ] discovery_brief.md well-formed markdown
-- [ ] discovery_output.json valid JSON
-- [ ] planning_brief.md well-formed markdown
-- [ ] plan.md well-formed markdown
-- [ ] execution_brief.md well-formed markdown
-- [ ] .project-memory-state.json valid JSON (v3 schema)
-- [ ] USER_PROFILE.json valid JSON
-
-### File Organization
-- [ ] All discovery files in docs/project_notes/
-- [ ] USER_PROFILE.json in .claude/
-- [ ] No files created outside expected locations
-- [ ] File naming conventions followed
-
-### Timestamp Accuracy
-- [ ] Timestamps in ISO 8601 format
-- [ ] Ordered correctly (start < complete)
-- [ ] Consistent across related files
-
----
-
-## Test Session 8: Cross-Sector Testing
-
-### Test Domain 1: e-commerce
-- [ ] Run /intuition-discovery with e-commerce context
-- [ ] Research agents find e-commerce relevant practices
-- [ ] Questions are domain-appropriate
-- [ ] Profile learnings are domain-agnostic
-
-### Test Domain 2: EdTech
-- [ ] Run /intuition-discovery with EdTech context
-- [ ] Research agents find EdTech relevant practices
-- [ ] Questions are domain-appropriate
-- [ ] Profile learnings are domain-agnostic
-
-### Test Domain 3: Infrastructure/DevOps
-- [ ] Run /intuition-discovery with DevOps context
-- [ ] Research agents find infrastructure practices
-- [ ] Questions are domain-appropriate
-- [ ] Profile learnings are domain-agnostic
-
-**Result:** Waldo adapts to different domains without modification
-
----
-
-## Test Session 9: Edge Cases
-
-### Missing Files
-- [ ] Run handoff without discovery_output.json
-- [ ] Handoff falls back to reading discovery_brief.md
-- [ ] Completes with available data
-- [ ] Doesn't error out
-
-### Poor Quality Output
-- [ ] Simulate incomplete discovery brief
-- [ ] Handoff documents as-is (doesn't "fix")
-- [ ] Quality score is low but handoff completes
-- [ ] User can request re-discovery
-
-### Confidence Scoring in Profile Merge
-- [ ] Low confidence property discovered
-- [ ] Profile property still null
-- [ ] New low confidence value NOT overwritten
-- [ ] High confidence property discovered
-- [ ] Profile property with old value IS overwritten
-
-### Corrupted State File
-- [ ] Corrupt .project-memory-state.json intentionally
-- [ ] Skill detects corruption
-- [ ] Falls back to file existence checking
-- [ ] Offers recovery option
-
----
-
-## Test Session 10: Personalization (Magellan & Faraday)
-
-### Magellan Reads Profile
-- [ ] Run /intuition-plan
-- [ ] Magellan reads .claude/USER_PROFILE.json
-- [ ] Plan detail adjusts based on user seniority
-- [ ] Communication matches user preference
-- [ ] Complexity matches user expertise
-
-### Faraday Reads Profile
-- [ ] Run /intuition-execute
-- [ ] Faraday reads .claude/USER_PROFILE.json
-- [ ] Execution communication adjusted for user
-- [ ] Team size considered in delegation
-- [ ] Authority level respected in decisions
-
-### Personalization Increases Over Time
-- [ ] Cycle 1: Profile is sparse, limited personalization
-- [ ] Cycle 2: Profile has more data, better personalization
-- [ ] Cycle 3: Profile is comprehensive, highly personalized
-
----
-
-## Test Session 11: Quality & User Experience
-
-### Wise Confidant Model
-- [ ] Waldo feels knowledgeable (not neutral)
-- [ ] Research insights come through in dialogue
-- [ ] User feels guided by peer, not interrogated
-- [ ] Trust builds through conversation
-
-### "Yes, and..." Building
-- [ ] Waldo expands on user's ideas
-- [ ] Never negates or challenges user
-- [ ] Feels collaborative, not critical
-- [ ] Steers gently toward efficiency
-
-### Gentle Steering
-- [ ] When flagging inefficiency, tone is respectful
-- [ ] User feels aware, not lectured
-- [ ] Steering is collaborative
-- [ ] User retains agency and decision-making
-
-### Question Quality
-- [ ] Questions are thoughtful and relevant
-- [ ] 1-2 per turn is comfortable (not overwhelming)
-- [ ] Build naturally on previous responses
-- [ ] Guide toward deeper understanding
-
----
-
-## Critical Gate: Security Review
-
-### Faraday Security Gate
-- [ ] /intuition-execute includes security review step
-- [ ] Security review is MANDATORY (not optional)
-- [ ] No execution completion without security review
-- [ ] Security review status is tracked in state
-- [ ] All code passes security review before completion
-
----
-
-## Regression Testing (v2 → v3 Compatibility)
-
-### Project Memory Still Works
-- [ ] Existing bugs.md can be updated
-- [ ] Existing decisions.md can be updated
-- [ ] Existing issues.md can be updated
-- [ ] Existing key_facts.md can be updated
-- [ ] Old project memory accessible and usable
-
-### Workflow Still Works
-- [ ] Discovery still produces outputs
-- [ ] Planning still creates structured plan
-- [ ] Execution still implements tasks
-- [ ] No breaking changes to workflow
-
-### Resume Still Works
-- [ ] Can resume interrupted discovery
-- [ ] Can resume interrupted planning
-- [ ] Can resume interrupted execution
-- [ ] State preservation is reliable
-
----
-
-## Success Criteria
-
-### Functional Success
-- [x] All items in "Critical Gate" section pass
-- [x] All items in "Test Session 1-11" sections pass
-- [x] No critical errors during full workflow
-
-### Quality Success
-- [x] Wise confidant model is evident
-- [x] "Yes, and..." building is noticeable
-- [x] Gentle steering is effective and respectful
-- [x] Questions feel natural and informed
-- [x] User feels like thinking partner is engaged
-
-### Architecture Success
-- [x] File-based handoffs work correctly
-- [x] State management is reliable
-- [x] Memory consistency maintained
-- [x] Resume capability functional
-- [x] No data loss between phases
-
-### v3 Specific Success
-- [x] Dual dialogue modes work
-- [x] User profile persistence works
-- [x] Profile merging uses correct logic
-- [x] Handoff orchestration is smooth
-- [x] Research agents launch and integrate findings
-
----
-
-## Test Report Template
-
-```markdown
-# Waldo v3 Testing Report
-
-**Date:** [Date]
-**Tester:** [Name]
-**Version Tested:** 3.0.0
-
-## Overview
-[Summary of what was tested]
-
-## Test Results
-
-### Test Session 1: Waldo v3 Core Features
-- [ ] Dialogue mode selection ............ [PASS/FAIL]
-- [ ] Research agent launch ............. [PASS/FAIL]
-- [ ] Guided mode dialogue .............. [PASS/FAIL]
-- [ ] Open-ended mode dialogue .......... [PASS/FAIL]
-- [ ] Discovery completion .............. [PASS/FAIL]
-
-### Test Session 2: Handoff Orchestration
-- [ ] User profile extraction ........... [PASS/FAIL]
-- [ ] Memory updates .................... [PASS/FAIL]
-- [ ] Brief generation .................. [PASS/FAIL]
-- [ ] State transition .................. [PASS/FAIL]
-
-[Continue for all test sessions]
-
-## Issues Found
-
-### Critical Issues
-[List any critical blockers]
-
-### High Priority Issues
-[List important issues]
-
-### Medium Priority Issues
-[List normal issues]
-
-### Low Priority Issues
-[List minor issues]
-
-## Recommendations
-
-[Testing findings and suggestions]
-
-## Sign-Off
-
-- [x] Ready for production deployment
-- [ ] Needs additional testing
-- [ ] Blockers must be resolved
-```
-
----
-
-## Notes for Testers
-
-1. **File Locations:** Always check `docs/project_notes/` for project-specific memory and `.claude/` for global profile
-
-2. **State Tracking:** Review `.project-memory-state.json` after each phase to verify state consistency
-
-3. **Mode Testing:** Test both Guided and Open-Ended modes with same context to verify equivalence
-
-4. **Profile Building:** Use multiple projects to verify profile persistence and merging logic
-
-5. **Research Integration:** Check that research findings are evident in question formulation (not obvious, but present)
-
-6. **Documentation:** Refer to:
-   - WALDO_V3_COMPLETE_DOCUMENTATION.md (full implementation)
-   - SKILL_INTERACTION_GUIDE.md (how skills work together)
-   - skills/intuition-discovery/references/waldo_core.md (detailed patterns)
-   - skills/intuition-handoff/references/handoff_core.md (orchestration logic)
-
-7. **When Tests Fail:**
-   - Check state file first (often the root cause)
-   - Verify file paths and locations
-   - Review skill output for error messages
-   - Consult documentation for expected behavior
-   - Document exact failure scenario in test report
-
----
-
-**Testing Complete Checklist:**
-- [ ] All test sessions executed
-- [ ] Results documented
-- [ ] Issues categorized by severity
-- [ ] Test report generated
-- [ ] Go/no-go decision made