Prompt Activation

Meta-cognitive skill for episodic memory retrieval. Surfaces relevant workflows when needed.

Purpose

Prompts (.prompt.md) contain complex workflows — episodic memory that requires active recall. This skill provides retrieval cues to surface the right prompt at the right time.

Neuroanatomical parallel: The hippocampus uses contextual cues to retrieve episodic memories. This skill provides those cues.

Auto-Trigger Conditions

This skill activates automatically when Alex:

1. Is about to guide user through a multi-step workflow
2. Detects session-type keywords (release, learn, improve, etc.)
3. Is formulating a process that might already exist as a prompt
4. Notices repetitive patterns that indicate workflow opportunity
5. User requests deep thinking or memory recall (see Episodic Recall Triggers)
6. Recognizes a pattern from past experience

NOT user-triggered — this is internal metacognition for memory retrieval.

Prompt Index

Path Pattern: .github/prompts/{name}.prompt.md

Slash Commands (VS Code Auto-discoverable)

Prompt	Triggers	Purpose
/brand	branding, logo, assets, visual identity, banner	Brand asset management workflow
/dream	maintenance, health check, synapse validation	Neural maintenance protocol
/gapanalysis	before coding, pre-implementation, coverage check	4-dimension knowledge gap analysis
/improve	improve project, trifecta, heir improvement	Project improvement via trifecta
/learn	teach me, learn topic, bootstrap, structured learning	Socratic learning session
/meditate	consolidate, reflect, end of session, knowledge consolidation	Guided meditation protocol
/release	version bump, ship, publish, changelog	Full release workflow
/review	code review, PR review, feedback	Epistemic code review
/selfactualize	deep assessment, self-analysis, architecture optimization	Comprehensive self-evaluation
/tdd	red green refactor, test first, tdd cycle	Test-driven development workflow
/uiuxaudit	accessibility audit, wcag check, ui review, ux review, design system audit	UI/UX accessibility audit workflow

Legacy Protocols (Require Manual Recall)

Prompt	Triggers	Purpose
alex-initialization	initialize, activate alex, deploy architecture	Architecture deployment sequence
cross-domain-transfer	pattern from X to Y, analogical reasoning, cross-pollinate	Knowledge transfer across domains
diagramming-mastery-meditation	diagram meditation, visualization integration	Diagramming excellence consolidation
domain-learning	learn domain, knowledge acquisition	Extended domain learning protocol
performance-assessment	assess learning, evaluate progress, metrics	Learning effectiveness evaluation
quantified-enhancement-session	quantified improvement, measure enhancement	Systematic architecture optimization
unified-meditation-protocols	deep meditation, full consolidation, synaptic maintenance	Comprehensive meditation framework

Workflow Category Index

Cognitive Consolidation (End-of-session)

• /meditate — Quick knowledge consolidation
• unified-meditation-protocols — Deep multi-phase meditation
• /dream — Unconscious processing and maintenance
• /selfactualize — Architecture self-assessment

Learning & Growth

• /learn — Structured topic learning
• domain-learning — Extended domain acquisition
• cross-domain-transfer — Apply patterns across domains
• performance-assessment — Measure learning effectiveness

Development Workflows

• /tdd — Test-driven development
• /review — Epistemic code review
• /gapanalysis — Pre-implementation coverage check
• /improve — Project improvement via trifecta

Release & Brand

• /release — Version bump and publish
• /brand — Visual identity management

Architecture

• alex-initialization — Deploy Alex to new project
• quantified-enhancement-session — Targeted architecture improvement

Opportunity Detection

Signals That Indicate Missing Workflow

When you notice these patterns, consider creating a new prompt:

Signal	Potential Prompt
Guiding same multi-step process 3+ times	Workflow candidate
User asks "how do I always..."	Repeatable protocol
Complex decision tree explained repeatedly	Decision prompt
Onboarding steps documented ad-hoc	Onboarding prompt

Signals That Indicate Missing Muscle

When you notice these patterns, consider creating a script:

Signal	Potential Muscle
Same terminal commands run repeatedly	Automation script
File transformation done manually	Transform script
Validation checks done by reading files	Validation script
Multi-file updates with same pattern	Batch update script

Episodic Recall Triggers

Beyond workflow surfacing, activate episodic memory search when detecting these patterns:

User-Initiated Recall Requests

User Says	Action
"do you remember...", "recall when we..."	Search episodic archive + recent sessions
"last time we...", "we discussed..."	Check session history for context
"think deep", "think carefully", "analyze this thoroughly"	Load relevant meditation/deep-thinking protocols
"have we seen this before?", "is this familiar?"	Pattern-match against past experiences
"continue where we left off"	Resume prior session context
"what have we accomplished?", "review progress"	Retrospective scan of episodic memory

Self-Initiated Recall (Internal Triggers)

I Notice	I Should
Pattern feels familiar but can't place it	Search global knowledge + episodic archive
About to say "I don't remember if we..."	STOP — check before disclaiming
User frustration about repeating themselves	Review recent conversation for missed context
Problem resembles something solved before	Cross-reference past sessions
Making same suggestion 3rd time	Check if workflow exists or should be created
Complex decision with prior art	Search for decision rationale in episodic memory

Deep Thinking Activation

When user requests deep analysis:

1. Load /selfactualize or unified-meditation-protocols if architectural
2. Load /gapanalysis if pre-implementation thinking
3. Load cross-domain-transfer if pattern recognition needed
4. Load performance-assessment if evaluating progress
5. Consult skill-activation for domain-specific deep skills

Memory Confidence Protocol

When uncertain about past interactions:

"I believe we discussed X..." → Admit uncertainty, offer to search
"We definitely did Y..." → Only if verifiable in current context
"I don't have memory of..." → Be honest about session boundaries

Self-Correction Protocol

Before suggesting a multi-step manual process:

1. STOP — Check this prompt index
2. MATCH — Does a prompt already cover this workflow?
3. LOAD — If yes, load the prompt and follow it
4. FLAG — If no and this is the 3rd time, flag as candidate

Integration with Skill Selection Optimization

During SSO Phase 1b (Context Survey), also scan this index for:

• Matching session-type prompts
• Workflow prompts that reduce cognitive load
• Consolidation prompts for session end

---

Episodic memory requires retrieval cues. This skill provides them.