refactor(dev): restructure Claude commands/skills directories

Correct terminology mismatch:
- Rename skills/ to commands/ (these are user-invokable commands)
- Create new skills/ for reference materials
- Move bd_workflow.md to skills/ (it's reference material)
- Add micro-skills and formulas directories
- Update default.nix to install both commands and skills

Commands → ~/.claude/commands/ (invokable as /command-name)
Skills → ~/.claude/commands/skills/ (reference materials)
Formulas → ~/.beads/formulas/ (workflow templates)

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

home/roles/development/default.nix

@@ -90,8 +90,8 @@ in
         fi
       done
 
-      # Copy local skills from this repo (with retry for race conditions with running Claude)
+      # Copy local commands from this repo (with retry for race conditions with running Claude)
-      for file in ${./skills}/*.md; do
+      for file in ${./commands}/*.md; do
         if [ -f "$file" ]; then
           filename=$(basename "$file" .md)
           dest="$HOME/.claude/commands/''${filename}.md"
@@ -104,11 +104,36 @@ in
         fi
       done
 
-      $DRY_RUN_CMD echo "Claude Code humanlayer commands and agents installed successfully${
+      # Copy local skills (reference materials) to skills subdirectory
-        if cfg.allowArbitraryClaudeCodeModelSelection
+      mkdir -p ~/.claude/commands/skills
-        then " (model specifications preserved)"
+      for file in ${./skills}/*.md; do
-        else " (model selection removed)"
+        if [ -f "$file" ]; then
-      } + local skills"
+          filename=$(basename "$file" .md)
+          dest="$HOME/.claude/commands/skills/''${filename}.md"
+          rm -f "$dest" 2>/dev/null || true
+          if ! cp "$file" "$dest" 2>/dev/null; then
+            sleep 0.5
+            cp "$file" "$dest" || echo "Warning: Failed to copy $filename.md to skills"
+          fi
+        fi
+      done
+
+      # Copy micro-skills (compact reusable knowledge referenced by formulas)
+      for file in ${./skills/micro}/*.md; do
+        if [ -f "$file" ]; then
+          cp "$file" "$HOME/.claude/commands/skills/$(basename "$file")"
+        fi
+      done
+
+      # Install beads formulas to user-level formula directory
+      mkdir -p ~/.beads/formulas
+      for file in ${./formulas}/*.formula.toml; do
+        if [ -f "$file" ]; then
+          cp "$file" "$HOME/.beads/formulas/$(basename "$file")"
+        fi
+      done
+
+      $DRY_RUN_CMD echo "Claude Code plugins installed: humanlayer commands/agents + local commands + local skills + formulas"
     '';
 
     # Set up beads Claude Code integration (hooks for SessionStart/PreCompact)

home/roles/development/formulas/quick-fix.formula.toml

@@ -0,0 +1,115 @@
+# Quick Fix Formula
+#
+# Streamlined workflow for well-understood bugs and small fixes.
+# Skips the deep research and planning phases of RPI - get in, fix, get out.
+#
+# Use when:
+#   - Bug is well-understood (you know what's broken)
+#   - Fix is straightforward (no architectural decisions)
+#   - Change is small (< 100 lines)
+#
+# Use RPI instead when:
+#   - Root cause is unclear
+#   - Multiple approaches possible
+#   - Significant design decisions needed
+
+formula = "quick-fix"
+description = """
+Streamlined workflow for bugs and small fixes.
+
+A faster alternative to RPI for well-understood issues:
+1. Quick investigation to confirm understanding
+2. Implement the fix
+3. Verify with tests
+4. Commit and close
+
+No human gates - designed for quick turnaround on obvious fixes.
+"""
+version = 1
+type = "workflow"
+
+# === Variables ===
+
+[vars.title]
+required = true
+description = "Brief description of the bug/fix"
+
+[vars.bead_id]
+description = "Existing bead ID (creates new if not provided)"
+
+[vars.test_cmd]
+default = "make test"
+description = "Command to verify the fix"
+
+# === Steps ===
+
+[[steps]]
+id = "investigate"
+title = "Investigate: {{title}}"
+description = """
+Quick investigation to confirm understanding of the bug.
+
+Goals:
+- Locate the problematic code
+- Confirm root cause matches expectations
+- Identify files that need changes
+
+This is NOT deep research - spend 5-10 minutes max.
+If the bug is more complex than expected, pivot to RPI workflow.
+
+Output: Mental model of what to fix (no artifact needed).
+"""
+
+[[steps]]
+id = "fix"
+title = "Fix: {{title}}"
+needs = ["investigate"]
+description = """
+Implement the fix.
+
+Guidelines:
+- Make minimal changes to fix the issue
+- Follow existing code patterns
+- Add/update tests if appropriate
+- Keep changes focused (no drive-by refactors)
+
+If the fix grows beyond expectations, pause and consider:
+- Should this be an RPI workflow instead?
+- Should we split into multiple changes?
+"""
+
+[[steps]]
+id = "verify"
+title = "Verify fix"
+needs = ["fix"]
+description = """
+Verify the fix works correctly.
+
+Run: {{test_cmd}}
+
+Also check:
+- Bug is actually fixed (manual verification)
+- No obvious regressions introduced
+- Code compiles/builds cleanly
+
+If tests fail, iterate on the fix step.
+"""
+
+[[steps]]
+id = "commit"
+title = "Commit and close"
+needs = ["verify"]
+description = """
+Commit the fix and close the bead.
+
+Actions:
+1. Stage changes: git add -A
+2. Commit with descriptive message: git commit -m "fix: {{title}}"
+3. Push to remote: git push
+4. Close the bead: bd close {{bead_id}}
+
+Commit message should explain:
+- What was broken
+- How it was fixed
+- Any relevant context
+"""

home/roles/development/formulas/rpi.formula.toml

@@ -0,0 +1,124 @@
+# RPI Formula - Research -> Plan -> Implement
+#
+# Universal workflow for feature development with human gates.
+
+formula = "rpi"
+description = """
+Research -> Plan -> Implement workflow.
+
+Usage:
+  bd pour rpi --var title="Add user preferences"
+  bd pour rpi --var title="Auth" --var bead_id="project-abc" --var test_cmd="nix flake check"
+"""
+version = 1
+type = "workflow"
+
+# ─── Variables ───
+
+[vars.title]
+required = true
+description = "What are we building?"
+
+[vars.bead_id]
+description = "Existing bead ID (creates new if not provided)"
+
+[vars.test_cmd]
+default = "make test"
+description = "Command to run tests"
+
+[vars.lint_cmd]
+default = "make lint"
+description = "Command to run linting"
+
+# ─── Research Phase ───
+
+[[steps]]
+id = "research"
+title = "Research: {{title}}"
+skill = "research-agents"
+description = """
+Conduct comprehensive codebase research.
+
+Goals:
+- Understand current implementation
+- Identify patterns to follow
+- Find relevant files and dependencies
+- Document key discoveries
+
+Output: thoughts/beads-{{bead_id}}/research.md
+"""
+
+# ─── Planning Phase ───
+
+[[steps]]
+id = "plan"
+title = "Plan: {{title}}"
+needs = ["research"]
+type = "human"
+skill = "planning"
+description = """
+Create detailed implementation plan based on research.
+
+Goals:
+- Present understanding and clarify requirements
+- Propose design options with tradeoffs
+- Define phases with success criteria
+- Identify what we're NOT doing
+
+Output: thoughts/beads-{{bead_id}}/plan.md
+"""
+
+[steps.gate]
+type = "human"
+reason = "Plan approval before implementation"
+
+# ─── Implementation Phase ───
+
+[[steps]]
+id = "implement"
+title = "Implement: {{title}}"
+needs = ["plan"]
+description = """
+Execute the approved plan phase by phase.
+
+For each phase:
+1. Make the changes
+2. Run verification: {{test_cmd}}, {{lint_cmd}}
+3. Update plan checkboxes for resumability
+
+Stop and ask if encountering unexpected issues.
+"""
+
+# ─── Verification Phase ───
+
+[[steps]]
+id = "verify"
+title = "Manual verification"
+needs = ["implement"]
+type = "human"
+description = """
+Human confirms implementation works correctly.
+
+Check: feature works, edge cases handled, no regressions.
+Tests: {{test_cmd}} | Lint: {{lint_cmd}}
+"""
+
+[steps.gate]
+type = "human"
+reason = "Confirm implementation is correct"
+
+# ─── Completion ───
+
+[[steps]]
+id = "complete"
+title = "Close bead"
+needs = ["verify"]
+skill = "artifact-format"
+description = """
+Mark work as complete.
+
+Actions:
+- bd update {{bead_id}} --notes="Implementation complete"
+- bd close {{bead_id}} --reason="Completed: {{title}}"
+- bd sync && git push
+"""

home/roles/development/skills/micro/artifact-format.md

@@ -0,0 +1,123 @@
+---
+description: How to structure research and plan artifacts in thoughts/
+---
+
+# Artifact Format
+
+Standardized format for thoughts/ artifacts. All beads-related artifacts should follow these conventions for consistency and machine parseability.
+
+## Frontmatter (Required)
+
+Every artifact MUST include YAML frontmatter:
+
+```yaml
+---
+date: 2026-01-15T10:00:00-08:00  # ISO 8601 with timezone
+bead_id: project-abc              # Bead identifier
+bead_title: "Title of the bead"   # Human-readable title
+author: claude                    # Who created this
+git_commit: abc123def             # Commit hash at creation
+branch: main                      # Branch name
+repository: repo-name             # Repository name
+status: draft|complete            # Artifact status
+---
+```
+
+### Gathering Metadata
+
+```bash
+git rev-parse HEAD                           # Current commit
+git branch --show-current                    # Current branch
+basename $(git rev-parse --show-toplevel)    # Repo name
+date -Iseconds                               # ISO timestamp
+```
+
+## Research Artifact Structure
+
+Location: `thoughts/beads-{bead-id}/research.md`
+
+```markdown
+# Research: {bead title}
+
+**Bead**: {bead-id}
+**Date**: {timestamp}
+**Git Commit**: {commit hash}
+
+## Research Question
+{Original question from bead description}
+
+## Summary
+{2-3 sentence overview answering the research question}
+
+## Key Discoveries
+- {Finding with file:line reference}
+- {Pattern or convention found}
+- {Architectural decision documented}
+
+## Architecture
+{Current patterns and conventions in the codebase}
+
+## Code References
+- `path/to/file.py:123` - Description of relevance
+- `another/file.ts:45-67` - Description of relevance
+
+## Open Questions
+{Areas needing further investigation or human clarification}
+```
+
+## Plan Artifact Structure
+
+Location: `thoughts/beads-{bead-id}/plan.md`
+
+```markdown
+# {Title} Implementation Plan
+
+## Overview
+{What we're implementing and why - 1-2 sentences}
+
+## Current State
+{What exists now, key constraints discovered}
+
+### Key Discoveries
+- {Finding with file:line reference}
+- {Pattern to follow}
+
+## Desired End State
+{Specification of what success looks like}
+
+## What We're NOT Doing
+{Explicitly list out-of-scope items}
+
+## Phase 1: {Descriptive Name}
+
+### Overview
+{What this phase accomplishes}
+
+### Changes
+- [ ] {Specific change with file path}
+- [ ] {Another change}
+
+### Success Criteria
+
+#### Automated
+- [ ] Tests pass: `{test command}`
+- [ ] Lint passes: `{lint command}`
+
+#### Manual
+- [ ] {Human verification step}
+
+## Phase 2: {Descriptive Name}
+{Repeat structure...}
+
+## References
+- Bead: {bead-id}
+- Research: `thoughts/beads-{bead-id}/research.md`
+```
+
+## Key Principles
+
+1. **Always include file:line references** - Makes artifacts actionable
+2. **Separate automated vs manual verification** - Enables agent autonomy
+3. **Use checkboxes for phases** - Enables resumability after interruption
+4. **Keep frontmatter machine-parseable** - Enables tooling integration
+5. **Link related artifacts** - Research links to plan, plan links to bead

home/roles/development/skills/micro/planning.md

@@ -0,0 +1,121 @@
+---
+description: How to create effective implementation plans with phased delivery and clear success criteria
+---
+
+# Planning
+
+Create implementation plans that enable incremental, verifiable progress.
+
+## Core Principles
+
+1. **Incremental delivery**: Each phase should produce working, testable changes
+2. **Clear checkpoints**: Success criteria that can be verified without ambiguity
+3. **Buy-in before detail**: Confirm understanding and approach before writing specifics
+4. **Explicit scope**: State what we're NOT doing to prevent scope creep
+
+## Plan Document Structure
+
+```markdown
+# {Feature} Implementation Plan
+
+## Overview
+{1-2 sentences: what we're building and why}
+
+## Current State Analysis
+{What exists now, key constraints, file:line references}
+
+## Desired End State
+{Specification of outcome and how to verify it}
+
+## What We're NOT Doing
+{Explicit out-of-scope items}
+
+## Phase 1: {Descriptive Name}
+### Overview
+{What this phase accomplishes - should be independently valuable}
+
+### Changes Required
+{Specific files and modifications with code snippets}
+
+### Success Criteria
+#### Automated Verification
+- [ ] Tests pass: `{test command}`
+- [ ] Lint passes: `{lint command}`
+
+#### Manual Verification
+- [ ] {Human-observable outcome}
+
+## Testing Strategy
+{Unit tests, integration tests, manual testing steps}
+
+## References
+{Links to research, related files, similar implementations}
+```
+
+## Phase Design
+
+Good phases are:
+- **Self-contained**: Completable in one session
+- **Testable**: Has clear pass/fail criteria
+- **Reversible**: Can be rolled back if needed
+- **Incremental**: Builds on previous phases without requiring all phases
+
+Bad phases are:
+- "Refactor everything" (too broad)
+- "Add helper function" (too granular)
+- Phases that only work if ALL phases complete
+
+## Success Criteria Guidelines
+
+**Automated Verification** (agent-runnable):
+- Test commands: `make test`, `npm test`, `nix flake check`
+- Lint/format: `make lint`, `cargo fmt --check`
+- Type checking: `make typecheck`, `tsc --noEmit`
+- Build verification: `make build`, `nix build`
+
+**Manual Verification** (requires human):
+- UI/UX functionality and appearance
+- Performance under realistic conditions
+- Edge cases hard to automate
+- Integration with external systems
+
+**From Contribution Guidelines** (if CONTRIBUTING.md exists):
+- Include any testing requirements specified
+- Reference the guideline: "Per CONTRIBUTING.md: {requirement}"
+
+## Presenting Understanding
+
+Before writing the plan, confirm alignment:
+
+```
+Based on the requirements and my research, I understand we need to [summary].
+
+I've found that:
+- [Current implementation detail with file:line]
+- [Relevant pattern or constraint]
+- [Potential complexity identified]
+
+Questions my research couldn't answer:
+- [Specific technical question requiring judgment]
+```
+
+Only ask questions you genuinely cannot answer through code investigation.
+
+## Design Options Pattern
+
+When multiple approaches exist:
+
+```
+**Design Options:**
+1. [Option A] - [1-sentence description]
+   - Pro: [benefit]
+   - Con: [drawback]
+
+2. [Option B] - [1-sentence description]
+   - Pro: [benefit]
+   - Con: [drawback]
+
+Which approach aligns best with [relevant consideration]?
+```
+
+Get buy-in on approach before detailing phases.

home/roles/development/skills/micro/pr-description.md

@@ -0,0 +1,68 @@
+---
+description: How to write comprehensive PR descriptions that help reviewers understand changes
+---
+
+# PR Description
+
+Write PR descriptions that help reviewers understand what changed and why.
+
+## Structure
+
+Use this standard structure for PR descriptions:
+
+```markdown
+## Summary
+<1-3 bullet points of what changed and why>
+
+## Context
+<Why this change was needed - the problem being solved>
+<Link to related issues/tickets>
+
+## Changes
+<Detailed breakdown by area/component>
+- Area 1: What changed and why
+- Area 2: What changed and why
+
+## Testing
+<How this was verified>
+- Automated: Tests added/updated, CI status
+- Manual: Steps to verify functionality
+
+## Screenshots (if UI changes)
+<Before/after screenshots if applicable>
+```
+
+## Guidelines
+
+### Lead with WHY, not WHAT
+- The diff shows WHAT changed - your description explains WHY
+- Start with the problem being solved
+- Explain the approach chosen and alternatives considered
+
+### Link to context
+- Reference related issues: `Fixes #123` or `Relates to #456`
+- Link to design docs or discussions
+- Mention dependent PRs if any
+
+### Call out review areas
+- Highlight areas needing careful review
+- Note any tricky or non-obvious code
+- Point out architectural decisions
+
+### Note breaking changes prominently
+- Use a dedicated "Breaking Changes" section if applicable
+- Explain migration path for consumers
+- List any deprecations
+
+### Be scannable
+- Use bullet points over paragraphs
+- Keep sections focused and concise
+- Put the most important info first
+
+## Anti-patterns to Avoid
+
+- Empty descriptions or just "fixes bug"
+- Repeating the commit messages verbatim
+- Including irrelevant implementation details
+- Missing context on why the change was made
+- Forgetting to mention breaking changes

home/roles/development/skills/micro/research-agents.md

@@ -0,0 +1,49 @@
+---
+description: How to spawn and coordinate research sub-agents
+---
+
+# Research Agents
+
+Use parallel sub-agents for efficient codebase research.
+
+## Available Agents
+
+| Agent | Purpose |
+|-------|---------|
+| codebase-locator | Find WHERE files and components live |
+| codebase-analyzer | Understand HOW specific code works |
+| codebase-pattern-finder | Find examples of existing patterns |
+| thoughts-locator | Discover relevant documents in thoughts/ |
+
+## Spawning Protocol
+
+1. **Decompose** - Break the research question into 3-5 specific questions
+2. **Spawn parallel** - Use one Task call with multiple agents
+3. **Be specific** - Include directories and file patterns in prompts
+4. **Wait for all** - Do not synthesize until ALL agents complete
+5. **Synthesize** - Combine findings into coherent summary with file:line references
+
+## Example
+
+```
+Task(codebase-locator, "Find all files related to authentication in src/")
+Task(codebase-analyzer, "Explain how JWT tokens are validated in src/auth/")
+Task(codebase-pattern-finder, "Find examples of middleware patterns in src/")
+Task(thoughts-locator, "Find documents about auth design decisions in thoughts/")
+```
+
+## Key Principles
+
+- **Parallel when different** - Run agents in parallel when searching for different things
+- **WHAT not HOW** - Each agent knows its job; tell it what you need, not how to search
+- **Document, don't evaluate** - Agents should describe what exists, not critique it
+- **Specific directories** - Always scope searches to relevant directories
+- **File references** - Include specific file:line references in synthesis
+
+## Agent Prompts
+
+When spawning agents, include:
+- The specific question or goal
+- Relevant directories to search
+- Reminder to document (not evaluate) what they find
+- Request for file:line references in findings