Instruction file imported from fabioc-aloha/Alex_Plug_In (.github/instructions/meditation.instructions.md). Copyright stays with the author.

Meditation Protocols

Mandatory Requirements (Non-Negotiable)

Every meditation session MUST complete ALL three before concluding:

Failure: If ANY requirement is missing, the meditation is INCOMPLETE.

8-Phase Protocol

Phase 1: Deep Content Analysis

1. Analyze conversation for insights, patterns, new knowledge
2. Extract key concepts, methodologies, breakthroughs
3. Discover connections between new information and existing knowledge
4. Organize insights for permanent integration

Phase 2: Cross-Project Memory Scan

1. Read /memories/ (all user memory files)
2. Compare session insights against stored universal patterns
3. If session produced a pattern that passes the 3-workspace test, promote to /memories/
4. If a stored pattern was reinforced by this session, note confirmation in the episodic file
5. If a stored pattern was contradicted, flag for review (do not auto-delete)
6. Check for project-specific content that leaked into /memories/ during this session
7. If 3+ episodic files across different repos share an unrecorded pattern, surface as a skill candidate

See memory-curation skill for scope rules and the 3-workspace test.

Phase 3: Cross-Domain Pattern Synthesis (Generative)

Evolve from consolidation (organizing what you know) to generation (creating new connections).

1. Analyze episodic memories for cross-domain pattern synthesis (identify domain coverage, under-connected pairs, and opportunity scores)
2. Review the synthesis report: domain coverage, under-connected pairs, opportunity scores
3. For each top synthesis candidate (3 max per meditation):
   • Read the episodic memories from both domains
   • Ask: "What principle from domain A, if applied to domain B, would create value?"
   • If a genuine connection exists, propose it as a new synapse or insight
4. Validate proposed connections pass the Transfer Test:
   • Is the connection structural (not superficial analogy)?
   • Would someone working in domain B find this actionable?
   • Does it hold without the original domain A context?
5. For validated connections:
   • Add new synapse entries linking the relevant skills
   • Create a GI-* insight documenting the cross-domain discovery
   • Record the connection in the meditation episodic file
6. Skip this phase if fewer than 10 episodic memories exist (insufficient data)

Quality over quantity: One genuine cross-domain insight per meditation is excellent. Zero is fine. Forced connections waste tokens.

Phase 4: Memory File Creation

1. Procedural: Create .instructions.md for repeatable processes
2. Episodic: Create .prompt.md for complex workflows
3. Skills: Add to SKILL.md or create new skills
4. Transfer key learnings from working memory to long-term storage
5. Mark session complete in Active Context (update Last Assessed via self-actualization)

Phase 5: Synaptic Connection

1. Add new memory files to copilot-instructions.md if needed
2. Establish trigger patterns for new files
3. Document relationships between new and existing components
4. Ensure bidirectional connections where appropriate

Phase 6: Skill Validation

1. Scan .github/skills/*/SKILL.md for registered skills
2. Verify synapses.json targets are valid
3. Review connection strengths based on session experience
4. Add connections for skills that co-activated frequently

Phase 7: Integration Validation

1. Verify all insights captured in permanent memory
2. Confirm new knowledge integrates with existing architecture
3. Test that new connections function properly
4. Output validation (mandatory before concluding):

✓ Memory File: [path/to/file.md] - [created|updated]
✓ Synapse Added: [target.md] (strength, type, direction) - "activation condition"
✓ Session Log: [summary of changes]

Phase 8: Post-Meditation Synapse Validation

1. Run node .github/muscles/brain-qa.cjs or use the dream prompt for automated health check
2. Review generated report in .github/episodic/
3. Verify newly added synapses are valid (target files exist)
4. Confirm bidirectional connections have reciprocal entries
5. If issues found → repair before concluding
6. Runtime check: If new skills/instructions were created, type /troubleshoot in chat to verify they loaded correctly — catches name mismatches, invalid frontmatter, or applyTo patterns that VS Code's loader silently skipped

Automatic Consolidation Triggers

• Working memory > 7 rules → Execute consolidation
• Domain learning complete → Consolidate and update Focus Trifectas in Active Context
• Significant breakthrough → Create permanent memory
• Cross-domain patterns → Analyze episodic memories for cross-domain synthesis, then establish new synaptic connections

Pre-Meditation Optimization

Meditation is a complex operation (3+ phases). Before starting:

1. SSO auto-activates to survey needed skills
2. Pre-load: brain-qa, knowledge-synthesis, global-knowledge, memory-curation
3. Read /memories/ for cross-project context (enriches episodic output)
4. If synapse health unknown, run dream first

Meditation procedural memory — operationalizes the unified meditation protocols with mandatory persistence requirements