initial: add all custom Claude.ai skills

bulletproof/templates/handoff.md

@@ -0,0 +1,24 @@
+# Handoff: [task] — [phase]
+**Date:** YYYY-MM-DD
+
+## Goal
+What we're building and why (1-2 sentences)
+
+## Approach
+Key architectural decisions (why X, not Y)
+
+## Done
+- [x] What's completed (with file references)
+- [ ] What remains
+
+## Current Problem / Next Step
+What to do next (specific and actionable)
+
+## Key Files
+Files the agent should read first:
+- `path/to/file1` — why
+- `path/to/file2` — why
+
+## Key Decisions Made
+- Decision 1: chose X because Y
+- Decision 2: chose A because B

bulletproof/templates/plan.md

@@ -0,0 +1,47 @@
+# Plan: [Name]
+**Spec:** specs/YYYY-MM-DD-<n>.md
+**Status:** planning | in_progress | completed
+
+## Challenge Log
+**Problem:** [1-2 sentences]
+**Chosen solution:** [approach]
+**Alternatives considered:**
+1. [Alternative A] — rejected because [reason]
+2. [Alternative B] — rejected because [reason]
+**Why chosen solution is better:** [justification]
+
+## Problems
+
+| # | Problem | Solution | Status |
+|---|---------|----------|--------|
+| 1 | ... | ... | pending |
+
+## Phases
+
+### Phase 1: [Name]
+- **Status:** pending | in_progress | completed
+- **Files:** [specific list]
+- **Changes:** [what exactly changes]
+- **TDD:** [which tests to write first]
+- **Gates:** ruff ✅ | pytest ✅ | compose config ✅ | pio build ✅ (as applicable)
+- **Impact:** [what other modules could be affected]
+- **Prompt for next session:**
+  ```
+  Read plans/YYYY-MM-DD-xxx.md, Phase 1.
+  Read spec at specs/YYYY-MM-DD-xxx.md.
+  Implement according to plan. Start with tests.
+  Do not modify files outside of: [list].
+  After completing:
+  1. Self-audit against spec (every acceptance criterion covered?)
+  2. Verify bugs are real (not cosmetic fixes)
+  3. Impact analysis (did we break anything else?)
+  4. Run all tier 1 gates
+  ```
+
+### Phase 2: [Name]
+...
+
+## Changelog
+
+| Date | Phase | Changes |
+|------|-------|---------|

bulletproof/templates/research.md

@@ -0,0 +1,30 @@
+# Research: [task name]
+**Date:** YYYY-MM-DD
+**Task size:** S | M | L
+
+## Current Architecture
+How it works now (key modules, data flow)
+
+## Affected Areas
+Which files/modules will change and why
+
+| # | File/Module | Why affected |
+|---|-------------|-------------|
+| 1 | ... | ... |
+
+## Codebase Patterns
+Conventions in use (style, state management, API patterns, testing approach)
+
+## Risks and Constraints
+What could break, edge cases, performance concerns
+
+## Open Questions
+What needs clarification from user or further research
+
+## Best Practices Found
+(if WebSearch was used — key findings and links)
+
+## Conclusion & Recommendation
+**Recommended approach:** [which solution is best and why]
+**Key reasons:** [2-3 concrete justifications]
+**Risks of this approach:** [what to watch for]

bulletproof/templates/spec.md

@@ -0,0 +1,35 @@
+# Spec: [Name]
+**Date:** YYYY-MM-DD
+**Author:** [who requested]
+
+## Problem
+What's broken or missing (1-3 sentences)
+
+## Goal
+What we want to achieve (measurable outcome)
+
+## Scope
+
+### In Scope
+- Concrete behaviors / user stories
+- ...
+
+### Out of Scope (not doing)
+- Clearly mark boundaries
+- ...
+
+## Acceptance Criteria
+- [ ] Criterion 1 (verifiable)
+- [ ] Criterion 2
+- [ ] Criterion 3
+- [ ] ...
+
+## Constraints
+- Technical limitations
+- Backward compatibility requirements
+- Performance requirements
+- Security requirements
+
+## Non-Goals
+- What is explicitly NOT a goal
+- ...