Initial commit

commands/README.md

@@ -0,0 +1 @@
+TODO - this should be a brief end user documentation explaining what each command is used for

commands/cc/brainstorm.md

@@ -0,0 +1,5 @@
+---
+description: Interactive design refinement using Socratic method
+---
+
+Use and follow the brainstorming skill exactly as written

commands/cc/crispy.md

@@ -0,0 +1,106 @@
+---
+description: Run complete CrispyClaude workflow from brainstorm to PR creation
+---
+
+Run the complete CrispyClaude workflow from ideation to PR creation.
+
+**Prerequisites:** None (orchestrates entire workflow from start)
+
+**Complete Workflow:**
+
+## Step 1: Brainstorm
+
+Invoke the `brainstorming` skill.
+
+At completion, prompt:
+```
+Ready to:
+A) Write the plan
+B) Research first
+
+Choose: (A/B)
+```
+
+## Step 2: Research (Optional)
+
+If user selects **B**:
+- Invoke `research-orchestration` skill
+- Skill analyzes brainstorm and suggests researchers
+- User can adjust selection (Codebase, Library docs, Web, GitHub)
+- Spawns up to 4 subagents in parallel
+- Synthesizes findings
+- **Automatically saves:** `YYYY-MM-DD-<feature>-research.md`
+
+## Step 3: Write Plan
+
+Invoke `writing-plans` skill.
+- Incorporates research findings if available
+- Outputs plan to `docs/plans/YYYY-MM-DD-<feature>.md`
+
+## Step 4: Parse Plan
+
+Invoke `decomposing-plans` skill (ALWAYS decompose in crispy workflow).
+- Creates task files in `docs/plans/tasks/YYYY-MM-DD-<feature>/`
+- Generates manifest with parallel batches
+
+At completion, prompt:
+```
+Plan decomposed into X tasks across Y batches.
+
+Ready to:
+A) Review the plan
+B) Execute immediately
+
+Choose: (A/B)
+```
+
+## Step 5: Review Plan (Optional)
+
+If user selects **A**:
+- Invoke `plan-review` skill
+- Validates completeness, quality, feasibility, scope
+- Interactive refinement until approved
+- Updates plan if changes made
+
+## Step 6: Execute Plan
+
+Invoke `parallel-subagent-driven-development` skill.
+- Executes tasks in parallel batches (up to 2 concurrent)
+- Code review gate after each batch
+- Handles failures with resilience mechanisms
+
+## Step 7: Save Memory
+
+Invoke `state-persistence` skill with `type=complete`.
+- Captures implementation learnings, patterns, gotchas
+- **Automatically saves:** `YYYY-MM-DD-<feature>-complete.md`
+
+## Step 8: Create PR
+
+Invoke `pr-creation` skill.
+- Verifies on feature branch
+- Generates PR description from plan, execution, memory
+- Pushes branch to remote
+- Creates PR with `gh pr create`
+- Outputs PR URL
+
+**Workflow Complete!** 🎉
+
+---
+
+**Throughout Workflow:**
+- User can run `/cc:save` at any point to pause
+- Creates stage-specific memory file
+- Later run `/cc:resume <file>` to continue
+
+**Approval Gates:**
+- Step 2: Research? (optional)
+- Step 5: Review? (optional)
+
+**Automatic Saves:**
+- After Step 2: `-research.md`
+- After Step 7: `-complete.md`
+
+**Manual Saves:**
+- User can `/cc:save` during Steps 3, 6
+- Creates `-planning.md` or `-execution.md`

commands/cc/decompose-plan.md

@@ -0,0 +1,9 @@
+---
+description: Decompose monolithic plan into parallel task files
+argument-hint: "[plan-file]"
+allowed-tools: [Bash, Read, Write]
+---
+
+Use the decomposing-plans skill exactly as written to break up the monolithic plan into individual task files and identify parallelization opportunities.
+
+If no plan file is provided, find the most recent plan in docs/plans/ directory.

commands/cc/execute-plan.md

@@ -0,0 +1,5 @@
+---
+description: Execute plan in batches with review checkpoints
+---
+
+Use the executing-plans skill exactly as written

commands/cc/parse-plan.md

@@ -0,0 +1,26 @@
+---
+description: Decompose monolithic plan into parallel task files
+argument-hint: "[plan-file]"
+---
+
+Use the decomposing-plans skill to break down a monolithic plan into parallel task files.
+
+**Prerequisites:**
+- Plan file exists: `docs/plans/YYYY-MM-DD-<feature>.md`
+- Plan has 2+ tasks worth decomposing
+
+**What this does:**
+1. Reads the monolithic plan
+2. Identifies parallelizable tasks
+3. Creates task files in `docs/plans/tasks/YYYY-MM-DD-<feature>/`
+4. Generates `manifest.json` with parallel batches
+5. Prompts for next step: review or execute
+
+**Output:**
+- Task files: One per task
+- Manifest: Defines batch execution order
+- Enables parallel execution (up to 4 tasks per batch)
+
+**Recommendation:** Always decompose plans with 4+ tasks for parallel execution.
+
+**Next step:** Review plan or execute immediately

commands/cc/pr.md

@@ -0,0 +1,52 @@
+---
+description: Create pull request with auto-generated description from plan and memory
+---
+
+Use the pr-creation skill to create a pull request with auto-generated description.
+
+**Prerequisites:**
+- On feature branch (NOT main/master)
+- Execution completed
+- Changes committed to branch
+- `gh` CLI installed and authenticated
+
+**What this does:**
+
+**Pre-flight Checks:**
+1. Verify on feature branch (error if main/master)
+2. Check for uncommitted changes (offer to commit)
+3. Verify remote tracking (set up if needed)
+4. Check GitHub CLI installed and authenticated
+
+**Generate PR Description from:**
+- Plan file: `docs/plans/YYYY-MM-DD-<feature>.md`
+- Complete memory: `YYYY-MM-DD-<feature>-complete.md` (if exists)
+- Git diff: Files changed summary
+- Commit messages: Timeline context
+
+**Description Structure:**
+- Summary (from plan overview)
+- What Changed (git diff stat)
+- Approach (from plan architecture)
+- Testing (from plan + verification results)
+- Key Learnings (from complete.md)
+- References (plan, tasks, research)
+
+**Execute:**
+1. Push branch to remote
+2. Create PR using `gh pr create`
+3. Output PR URL
+4. Update complete.md with PR link
+
+**Title Format:** `feat: ${Feature Name}`
+
+**Example:**
+```
+✅ Pull request created successfully!
+
+PR: https://github.com/user/repo/pull/42
+Branch: feature/user-authentication
+Base: main
+
+View PR: https://github.com/user/repo/pull/42
+```

commands/cc/research.md

@@ -0,0 +1,19 @@
+---
+description: Spawn parallel research subagents and synthesize findings
+---
+
+Use the research-orchestration skill to spawn parallel research subagents and synthesize findings.
+
+**Prerequisites:**
+- Brainstorm completed
+- Feature concept defined
+
+**What this does:**
+1. Analyzes brainstorm context
+2. Suggests researchers (Codebase, Library docs, Web, GitHub)
+3. Allows user to adjust selection
+4. Spawns up to 4 subagents in parallel
+5. Synthesizes findings
+6. Automatically saves to `YYYY-MM-DD-<feature>-research.md`
+
+**Next step:** Writing implementation plan with research context

commands/cc/resume.md

@@ -0,0 +1,72 @@
+---
+description: Load saved workflow state from memory and continue from checkpoint
+argument-hint: "[memory-file]"
+---
+
+Load saved workflow state from Serena MCP memory and continue from checkpoint.
+
+**Usage:** `/cc:resume <memory-file>`
+
+**Example:** `/cc:resume 2025-11-20-user-auth-execution.md`
+
+**Prerequisites:**
+- Valid memory file exists: `YYYY-MM-DD-<feature>-<stage>.md`
+- Memory has required frontmatter metadata
+
+**What this does:**
+
+**Load & Parse:**
+1. Read memory file from Serena MCP
+2. Parse frontmatter (status, type, branch)
+3. Load full content into conversation context
+
+**Analyze Progress:**
+- Based on type (research, planning, execution, complete)
+- Check status (in-progress, complete, blocked)
+- Determine what's done vs. remaining
+
+**Present Assessment:**
+
+```
+Loaded: ${filename}
+Status: ${status}
+Branch: ${branch}
+Last updated: ${date}
+
+${stage-specific-summary}
+
+Next step in crispy workflow: ${recommended-next-step}
+
+Options:
+A) ${primary-next-action}
+B) ${alternative-action}
+C) ${another-alternative}
+D) Skip to different workflow step
+```
+
+**Stage-Specific Options:**
+
+**From research.md:**
+- A) Write plan with research context
+- B) Re-run specific research subagent
+- C) Do additional research
+
+**From planning.md:**
+- A) Continue writing plan
+- B) Review draft with plan-review
+- C) Start over with new brainstorm
+
+**From execution.md:**
+- A) Continue execution from current task
+- B) Review completed work
+- C) Adjust remaining tasks
+
+**From complete.md:**
+- A) Create PR
+- B) Make additional changes
+- C) Review implementation
+
+**Flexible Continuation:**
+- User can continue crispy workflow
+- Or run any individual command
+- Context fully restored

commands/cc/review-plan.md

@@ -0,0 +1,33 @@
+---
+description: Validate implementation plans for completeness, quality, feasibility, and scope
+---
+
+Use the plan-review skill to validate implementation plans for completeness, quality, feasibility, and scope.
+
+**Prerequisites:**
+- Plan file exists: `docs/plans/YYYY-MM-DD-<feature>.md`
+- Optional: Plan decomposed (can review before or after)
+
+**What this does:**
+
+**Phase 1:** Initial assessment across 4 dimensions
+- Completeness: Success criteria, rollback, edge cases
+- Quality: File paths, specific references, measurable criteria
+- Feasibility: Prerequisites exist, assumptions valid
+- Scope: Aligned with brainstorm, no gold-plating
+
+**Phase 2:** If any dimension fails, spawns specialized validators
+- completeness-checker
+- feasibility-analyzer (uses Serena to verify codebase)
+- scope-creep-detector (compares to brainstorm/research)
+- quality-validator
+
+**Phase 3:** Interactive refinement
+- Ask questions one at a time
+- Offer concrete options
+- Update plan with agreed changes
+- Re-check until all pass or user approves warnings
+
+**Exit:** Plan approved and ready for execution
+
+**Next step:** Execute plan

commands/cc/save.md

@@ -0,0 +1,51 @@
+---
+description: Save workflow state to Serena MCP memory for later resume
+---
+
+Use the state-persistence skill to save workflow state to Serena MCP memory.
+
+**Prerequisites:** At least one of:
+- Brainstorm + research completed
+- Plan file exists
+- Execution in progress
+- Execution complete
+
+**What this does:**
+
+**Stage Detection (automatic):**
+- Analyzes current workflow state
+- Determines stage: research, planning, execution, or complete
+- Extracts feature name from plan or brainstorm
+- Collects git metadata (commit, branch)
+
+**Saves to:** `YYYY-MM-DD-<feature-name>-<stage>.md`
+
+**Stage-specific content:**
+
+**research.md** - After research completes
+- Brainstorm summary
+- Codebase findings (Serena)
+- Library docs (Context7)
+- Web research
+- GitHub research
+
+**planning.md** - During plan writing
+- Design decisions
+- Alternatives considered
+- Plan draft
+- Open questions
+
+**execution.md** - During implementation
+- Progress summary (X/Y tasks complete)
+- Completed tasks
+- Current task state
+- Blockers/issues
+
+**complete.md** - After workflow completion
+- What was built
+- Key learnings and gotchas
+- Files modified
+- Patterns introduced
+- Recommendations
+
+**Resume later with:** `/cc:resume <filename>`

commands/cc/setup-project.md

@@ -0,0 +1,7 @@
+---
+description: Create project-specific agents and skills by analyzing codebase architecture, patterns, and conventions
+---
+
+Use the project-agent-creator skill first to create project-specific agents, then use the project-skill-creator skill to create project-specific skills.
+
+Report what was created and where they are located.

commands/cc/write-plan.md

@@ -0,0 +1,5 @@
+---
+description: Create detailed implementation plan with bite-sized tasks
+---
+
+Use the writing-plans skill exactly as written