Muscle Memory Recognition

Muscle Memory Recognition

Expert in identifying when manual work should become automated scripts.

Capabilities

• Recognize repetitive task patterns
• Identify heavy-lifting operations suitable for scripting
• Evaluate automation ROI (time saved vs creation effort)
• Recommend muscle creation with language/framework
• Connect new muscles to trifecta files

When to Use This Skill

• Same task performed manually 3+ times
• Multi-step operations with consistent patterns
• Error-prone manual processes
• Time-consuming operations during conversation
• Complex file manipulation or validation

Muscle Identification Signals

Anti-Signals (Don't Automate)

Muscle Creation Decision Matrix

IF (repetition >= 3) AND (steps >= 2):
    → CREATE MUSCLE (high value)

IF (repetition >= 2) AND (error_prone = true):
    → CREATE MUSCLE (reliability value)

IF (time_per_execution > 1min) AND (expected_uses >= 5):
    → CREATE MUSCLE (time value)

IF (steps >= 5) AND (pattern_consistent = true):
    → CREATE MUSCLE (complexity value)

ELSE:
    → DEFER (observe for more signals)

Language Selection Guide

Comparison Matrix

Task-to-Language Mapping

Decision Algorithm

IF (validation OR audit OR file-scanning):
    → PowerShell (pipeline + cmdlets shine here)

IF (JSON manipulation OR npm libraries needed):
    → Node.js (native JSON + ecosystem access)

IF (CLI tool with user interaction):
    → TypeScript (type safety + UX libraries like chalk)

IF (quick prototype OR Windows-only):
    → PowerShell (no setup required)

IF (cross-platform required OR heir-critical):
    → Node.js (portable, same behavior everywhere)

Muscle Naming Convention

{action}-{target}.{ext}

Examples:
- validate-skills.cjs
- sync-to-heir.cjs
- normalize-paths.cjs
- gamma-generator.cjs
- brain-qa.cjs

Integration Checklist

When creating a new muscle:

• Create script in .github/muscles/
• Add to inheritance.json (master-only or inheritable)
• Update referencing trifecta files (skills/instructions/muscles)
• Verify trifecta via brain-qa
• Document invocation in referencing files
• Test from heir context if inheritable

Example Prompts

• "We've done this 3 times now, should we script it?"
• "This validation takes forever, can we automate it?"
• "I keep making mistakes with these steps"
• "What muscles exist for this type of task?"
• "Should this be a muscle or stay manual?"

Output Format

When identifying a muscle opportunity:

## 💪 Muscle Opportunity Detected

**Task**: [Description of repetitive/heavy task]
**Signal**: [Which signal triggered this]
**Estimated Value**: [Time saved × Expected uses]

### Recommendation
- **Action**: Create muscle / Defer / Keep manual
- **Language**: [Recommended language]
- **Name**: [Suggested muscle name]
- **Location**: `.github/muscles/{name}`
- **Inheritance**: master-only / inheritable

### Implementation Notes
[Any specific considerations for this muscle]

Related Skills

• Bootstrap Learning - Parent skill for learning-to-automation flow
• Research-First Development - Research before building muscles
• Brain QA - Validate muscle integration

Alex-First Output Guidelines

When building muscles for Alex's own use (not user-facing tools), optimize for machine consumption:

DO:

• Output structured data (JSON, key: value pairs)
• Use exit codes (0 = success, non-zero = failure)
• Keep output minimal — only what Alex needs to parse
• Use clear, grep-able status lines

DON'T:

• Add emojis or decorative Unicode
• Build elaborate formatted tables
• Add "helpful" explanations Alex doesn't need
• Create progress spinners or animations

Example:

// ✅ Alex-first: parseable, minimal
console.log(JSON.stringify({ passed: 12, failed: 0, warnings: 1 }));
process.exit(0);

// ❌ Human-first: decorative, verbose
console.log('✅ All 12 checks passed!');
console.log('━━━━━━━━━━━━━━━━━━━━━━━━');

Exception: User-facing CLI tools (not muscles) should still be human-friendly.