renn-studio 0.6.0

package/commands/renn.run-stage.md

@@ -0,0 +1,343 @@
+---
+name: renn.run-stage
+description: Execute all runs in a stage with batch-based parallelization
+argument-hint: "<stage-number> [--gaps-only]"
+allowed-tools:
+  - Read
+  - Write
+  - Edit
+  - Glob
+  - Grep
+  - Bash
+  - Task
+  - TodoWrite
+  - AskUserQuestion
+---
+
+<objective>
+Execute all runs in a stage using batch-based parallel execution.
+
+Orchestrator stays lean: discover runs, analyze dependencies, group into batches, spawn subagents, collect results. Each subagent loads the full run-plan context and handles its own run.
+
+Context budget: ~15% orchestrator, 100% fresh per subagent.
+</objective>
+
+<execution_context>
+@~/.claude/renn/references/ui-brand.md
+@~/.claude/renn/workflows/run-stage.md
+</execution_context>
+
+<context>
+Stage: $ARGUMENTS
+
+**Flags:**
+- `--gaps-only` — Execute only gap closure runs (runs with `gap_closure: true` in frontmatter). Use after audit creates fix runs.
+
+@.renn/track.md
+@.renn/pulse.md
+</context>
+
+<process>
+0. **Resolve Horsepower Profile**
+
+   Read horsepower profile for agent spawning:
+   ```bash
+   HORSEPOWER=$(cat .renn/config.json 2>/dev/null | grep -o '"horsepower"[[:space:]]*:[[:space:]]*"[^"]*"' | grep -o '"[^"]*"$' | tr -d '"' || echo "balanced")
+   ```
+
+   Default to "balanced" if not set.
+
+   **Model lookup table:**
+
+   | Agent | max | balanced | eco |
+   |-------|-----|----------|-----|
+   | renn-runner | opus | sonnet | sonnet |
+   | renn-auditor | sonnet | sonnet | haiku |
+
+   Store resolved models for use in Task calls below.
+
+1. **Validate stage exists**
+   - Find stage directory matching argument
+   - Count run.md files
+   - Error if no runs found
+
+2. **Discover runs**
+   - List all *-run.md files in stage directory
+   - Check which have *-recap.md (already complete)
+   - If `--gaps-only`: filter to only runs with `gap_closure: true`
+   - Build list of incomplete runs
+
+3. **Group by batch**
+   - Read `batch` from each run's frontmatter
+   - Group runs by batch number
+   - Report batch structure to user
+
+4. **Execute batches**
+   For each batch in order:
+   - Spawn `renn-runner` for each run in batch (parallel Task calls)
+   - Wait for completion (Task blocks)
+   - Verify RECAPs created
+   - Proceed to next batch
+
+5. **Aggregate results**
+   - Collect recaps from all runs
+   - Report stage completion status
+
+6. **Commit any orchestrator corrections**
+   Check for uncommitted changes before verification:
+   ```bash
+   git status --porcelain
+   ```
+
+   **If changes exist:** Orchestrator made corrections between runner completions. Stage and commit them individually:
+   ```bash
+   # Stage each modified file individually (never use git add -u, git add ., or git add -A)
+   git status --porcelain | grep '^ M' | cut -c4- | while read file; do
+     git add "$file"
+   done
+   git commit -m "fix({stage}): orchestrator corrections"
+   ```
+
+   **If clean:** Continue to verification.
+
+7. **Verify stage goal**
+   Check config: `CHECKS_AUDITOR=$(cat .renn/config.json 2>/dev/null | grep -o '"auditor"[[:space:]]*:[[:space:]]*[^,}]*' | grep -o 'true\|false' || echo "true")`
+
+   **If `checks.auditor` is `false`:** Skip to step 8 (treat as passed).
+
+   **Otherwise:**
+   - Spawn `renn-auditor` subagent with stage directory and goal
+   - Auditor checks must_haves against actual codebase (not RECAP claims)
+   - Creates proof.md with detailed report
+   - Route by status:
+     - `passed` → continue to step 8
+     - `human_needed` → present items, get approval or feedback
+     - `gaps_found` → present gaps, offer `/renn.plan-stage {X} --gaps`
+
+8. **Update track and pulse**
+   - Update track.md, pulse.md
+
+9. **Update specs**
+   Mark stage requirements as Complete:
+   - Read track.md, find this stage's `Requirements:` line (e.g., "AUTH-01, AUTH-02")
+   - Read specs.md traceability table
+   - For each REQ-ID in this stage: change Status from "Pending" to "Complete"
+   - Write updated specs.md
+   - Skip if: specs.md doesn't exist, or stage has no Requirements line
+
+10. **Commit stage completion**
+    Check `COMMIT_PLANNING_DOCS` from config.json (default: true).
+    If false: Skip git operations for .renn/ files.
+    If true: Bundle all stage metadata updates in one commit:
+    - Stage: `git add .renn/track.md .renn/pulse.md`
+    - Stage specs.md if updated: `git add .renn/specs.md`
+    - Commit: `docs({stage}): complete {stage-name} stage`
+
+11. **Offer next steps**
+    - Route to next action (see `<offer_next>`)
+</process>
+
+<offer_next>
+Output this markdown directly (not as a code block). Route based on status:
+
+| Status | Route |
+|--------|-------|
+| `gaps_found` | Route C (gap closure) |
+| `human_needed` | Present checklist, then re-route based on approval |
+| `passed` + more stages | Route A (next stage) |
+| `passed` + last stage | Route B (milestone complete) |
+
+---
+
+**Route A: Stage verified, more stages remain**
+
+━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
+ RENN ► STAGE {Z} COMPLETE ✓
+━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
+
+**Stage {Z}: {Name}**
+
+{Y} runs executed
+Goal verified ✓
+
+───────────────────────────────────────────────────────────────
+
+## ▶ Next Up
+
+**Stage {Z+1}: {Name}** — {Goal from track.md}
+
+/renn.discuss-stage {Z+1} — gather context and clarify approach
+
+<sub>/clear first → fresh context window</sub>
+
+───────────────────────────────────────────────────────────────
+
+**Also available:**
+- /renn.plan-stage {Z+1} — skip discussion, plan directly
+- /renn.audit {Z} — manual acceptance testing before continuing
+
+───────────────────────────────────────────────────────────────
+
+---
+
+**Route B: Stage verified, milestone complete**
+
+━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
+ RENN ► MILESTONE COMPLETE 🎉
+━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
+
+**v1.0**
+
+{N} stages completed
+All stage goals verified ✓
+
+───────────────────────────────────────────────────────────────
+
+## ▶ Next Up
+
+**Audit milestone** — verify requirements, cross-stage integration, E2E flows
+
+/renn.audit-milestone
+
+<sub>/clear first → fresh context window</sub>
+
+───────────────────────────────────────────────────────────────
+
+**Also available:**
+- /renn.audit — manual acceptance testing
+- /renn.complete-milestone — skip audit, archive directly
+
+───────────────────────────────────────────────────────────────
+
+---
+
+**Route C: Gaps found — need additional planning**
+
+━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
+ RENN ► STAGE {Z} GAPS FOUND ⚠
+━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
+
+**Stage {Z}: {Name}**
+
+Score: {N}/{M} must-haves verified
+Report: .renn/stages/{stage_dir}/{stage}-proof.md
+
+### What's Missing
+
+{Extract gap summaries from proof.md}
+
+───────────────────────────────────────────────────────────────
+
+## ▶ Next Up
+
+**Plan gap closure** — create additional runs to complete the stage
+
+/renn.plan-stage {Z} --gaps
+
+<sub>/clear first → fresh context window</sub>
+
+───────────────────────────────────────────────────────────────
+
+**Also available:**
+- cat .renn/stages/{stage_dir}/{stage}-proof.md — see full report
+- /renn.audit {Z} — manual testing before planning
+
+───────────────────────────────────────────────────────────────
+
+---
+
+After user runs /renn.plan-stage {Z} --gaps:
+1. Architect reads proof.md gaps
+2. Creates runs 04, 05, etc. to close gaps
+3. User runs /renn.run-stage {Z} again
+4. Run-stage runs incomplete runs (04, 05...)
+5. Auditor runs again → loop until passed
+</offer_next>
+
+<batch_execution>
+**Parallel spawning:**
+
+Before spawning, read file contents. The `@` syntax does not work across Task() boundaries.
+
+```bash
+# Read each run and pulse.md
+RUN_01_CONTENT=$(cat "{run_01_path}")
+RUN_02_CONTENT=$(cat "{run_02_path}")
+RUN_03_CONTENT=$(cat "{run_03_path}")
+PULSE_CONTENT=$(cat .renn/pulse.md)
+```
+
+Spawn all runs in a batch with a single message containing multiple Task calls, with inlined content:
+
+```
+Task(prompt="Execute run at {run_01_path}\n\nRun:\n{run_01_content}\n\nProject pulse:\n{pulse_content}", subagent_type="renn-runner", model="{runner_model}")
+Task(prompt="Execute run at {run_02_path}\n\nRun:\n{run_02_content}\n\nProject pulse:\n{pulse_content}", subagent_type="renn-runner", model="{runner_model}")
+Task(prompt="Execute run at {run_03_path}\n\nRun:\n{run_03_content}\n\nProject pulse:\n{pulse_content}", subagent_type="renn-runner", model="{runner_model}")
+```
+
+All three run in parallel. Task tool blocks until all complete.
+
+**No polling.** No background agents. No TaskOutput loops.
+</batch_execution>
+
+<gate_handling>
+Runs with `autonomous: false` have gates. The run-stage.md workflow handles the full gate flow:
+- Subagent pauses at gate, returns structured state
+- Orchestrator presents to user, collects response
+- Spawns fresh continuation agent (not resume)
+
+See `@~/.claude/renn/workflows/run-stage.md` step `gate_handling` for complete details.
+</gate_handling>
+
+<drift_rules>
+During execution, handle discoveries automatically:
+
+1. **Auto-fix bugs** - Fix immediately, document in Recap
+2. **Auto-add critical** - Security/correctness gaps, add and document
+3. **Auto-fix blockers** - Can't proceed without fix, do it and document
+4. **Ask about architectural** - Major structural changes, stop and ask user
+
+Only rule 4 requires user intervention.
+</drift_rules>
+
+<commit_rules>
+**Per-Task Commits:**
+
+After each task completes:
+1. Stage only files modified by that task
+2. Commit with format: `{type}({stage}.{run}): {task-name}`
+3. Types: feat, fix, test, refactor, perf, chore
+4. Record commit hash for recap.md
+
+**Run Metadata Commit:**
+
+After all tasks in a run complete:
+1. Stage run artifacts only: run.md, recap.md
+2. Commit with format: `docs({stage}.{run}): complete [run-name] run`
+3. NO code files (already committed per-task)
+
+**Stage Completion Commit:**
+
+After all runs in stage complete (step 7):
+1. Stage: track.md, pulse.md, specs.md (if updated), proof.md
+2. Commit with format: `docs({stage}): complete {stage-name} stage`
+3. Bundles all stage-level state updates in one commit
+
+**NEVER use:**
+- `git add .`
+- `git add -A`
+- `git add src/` or any broad directory
+
+**Always stage files individually.**
+</commit_rules>
+
+<success_criteria>
+- [ ] All incomplete runs in stage executed
+- [ ] Each run has recap.md
+- [ ] Stage goal verified (must_haves checked against codebase)
+- [ ] proof.md created in stage directory
+- [ ] pulse.md reflects stage completion
+- [ ] track.md updated
+- [ ] specs.md updated (stage requirements marked Complete)
+- [ ] User informed of next steps
+</success_criteria>

package/commands/renn.set-profile.md

@@ -0,0 +1,116 @@
+---
+name: renn.set-profile
+description: Switch horsepower profile for RENN agents (max/balanced/eco)
+allowed-tools:
+  - Read
+  - Write
+  - Bash
+arguments:
+  - name: profile
+    description: "Profile name: max, balanced, or eco"
+    required: true
+---
+
+<objective>
+Switch the horsepower profile used by RENN agents. This controls which Claude model each agent uses, balancing quality vs token spend.
+</objective>
+
+<profiles>
+| Profile | Description |
+|---------|-------------|
+| **max** | Opus everywhere except read-only auditing |
+| **balanced** | Opus for planning, Sonnet for execution/auditing (default) |
+| **eco** | Sonnet for writing, Haiku for research/auditing |
+</profiles>
+
+<process>
+
+## 1. Validate argument
+
+```
+if $ARGUMENTS.profile not in ["max", "balanced", "eco"]:
+  Error: Invalid profile "$ARGUMENTS.profile"
+  Valid profiles: max, balanced, eco
+  STOP
+```
+
+## 2. Check for project
+
+```bash
+ls .renn/config.json 2>/dev/null
+```
+
+If no `.renn/` directory:
+```
+Error: No RENN project found.
+Run /renn.start first to initialize a project.
+```
+
+## 3. Update config.json
+
+Read current config:
+```bash
+cat .renn/config.json
+```
+
+Update `horsepower` field (or add if missing):
+```json
+{
+  "horsepower": "$ARGUMENTS.profile"
+}
+```
+
+Write updated config back to `.renn/config.json`.
+
+## 4. Confirm
+
+```
+✓ Horsepower profile set to: $ARGUMENTS.profile
+
+Agents will now use:
+[Show table from horsepower-profiles.md for selected profile]
+
+Next spawned agents will use the new profile.
+```
+
+</process>
+
+<examples>
+
+**Switch to eco mode:**
+```
+/renn.set-profile eco
+
+✓ Horsepower profile set to: eco
+
+Agents will now use:
+| Agent | Model |
+|-------|-------|
+| renn-architect | sonnet |
+| renn-runner | sonnet |
+| renn-auditor | haiku |
+| ... | ... |
+```
+
+**Switch to max mode:**
+```
+/renn.set-profile max
+
+✓ Horsepower profile set to: max
+
+Agents will now use:
+| Agent | Model |
+|-------|-------|
+| renn-architect | opus |
+| renn-runner | opus |
+| renn-auditor | sonnet |
+| ... | ... |
+```
+
+</examples>
+
+<success_criteria>
+- [ ] Profile validated as max, balanced, or eco
+- [ ] config.json updated with new horsepower value
+- [ ] Confirmation with model table displayed to user
+</success_criteria>

package/commands/renn.settings.md

@@ -0,0 +1,151 @@
+---
+name: renn.settings
+description: Configure RENN workflow toggles and horsepower profile
+allowed-tools:
+  - Read
+  - Write
+  - AskUserQuestion
+---
+
+<objective>
+Allow users to toggle workflow agents on/off and select horsepower profile via interactive settings.
+
+Updates `.renn/config.json` with workflow preferences and horsepower profile selection.
+</objective>
+
+<process>
+
+## 1. Validate Environment
+
+```bash
+ls .renn/config.json 2>/dev/null
+```
+
+**If not found:** Error - run `/renn.start` first.
+
+## 2. Read Current Config
+
+```bash
+cat .renn/config.json
+```
+
+Parse current values (default to `true` if not present):
+- `checks.research` — spawn scout during plan-stage
+- `checks.review` — spawn plan reviewer during plan-stage
+- `checks.auditor` — spawn auditor during run-stage
+- `horsepower` — which model each agent uses (default: `balanced`)
+- `git.branching_strategy` — branching approach (default: `"none"`)
+
+## 3. Present Settings
+
+Use AskUserQuestion with current values shown:
+
+```
+AskUserQuestion([
+  {
+    question: "Which horsepower profile for agents?",
+    header: "Horsepower",
+    multiSelect: false,
+    options: [
+      { label: "Max", description: "Opus everywhere except auditing (highest cost)" },
+      { label: "Balanced (Recommended)", description: "Opus for planning, Sonnet for execution/auditing" },
+      { label: "Eco", description: "Sonnet for writing, Haiku for research/auditing (lowest cost)" }
+    ]
+  },
+  {
+    question: "Spawn Stage Scout? (researches domain before planning)",
+    header: "Research",
+    multiSelect: false,
+    options: [
+      { label: "Yes", description: "Research stage goals before planning" },
+      { label: "No", description: "Skip research, plan directly" }
+    ]
+  },
+  {
+    question: "Spawn Plan Reviewer? (verifies runs before execution)",
+    header: "Review",
+    multiSelect: false,
+    options: [
+      { label: "Yes", description: "Verify runs meet stage goals" },
+      { label: "No", description: "Skip run verification" }
+    ]
+  },
+  {
+    question: "Spawn Auditor? (verifies stage completion)",
+    header: "Auditor",
+    multiSelect: false,
+    options: [
+      { label: "Yes", description: "Verify must-haves after execution" },
+      { label: "No", description: "Skip post-execution auditing" }
+    ]
+  },
+  {
+    question: "Git branching strategy?",
+    header: "Branching",
+    multiSelect: false,
+    options: [
+      { label: "None (Recommended)", description: "Commit directly to current branch" },
+      { label: "Per Stage", description: "Create branch for each stage (renn/stage-{N}-{name})" },
+      { label: "Per Milestone", description: "Create branch for entire milestone (renn/{version}-{name})" }
+    ]
+  }
+])
+```
+
+**Pre-select based on current config values.**
+
+## 4. Update Config
+
+Merge new settings into existing config.json:
+
+```json
+{
+  ...existing_config,
+  "horsepower": "max" | "balanced" | "eco",
+  "checks": {
+    "research": true/false,
+    "review": true/false,
+    "auditor": true/false
+  },
+  "git": {
+    "branching_strategy": "none" | "stage" | "milestone"
+  }
+}
+```
+
+Write updated config to `.renn/config.json`.
+
+## 5. Confirm Changes
+
+Display:
+
+```
+━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
+ RENN ► SETTINGS UPDATED
+━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
+
+| Setting              | Value |
+|----------------------|-------|
+| Horsepower Profile   | {max/balanced/eco} |
+| Stage Scout          | {On/Off} |
+| Plan Reviewer        | {On/Off} |
+| Auditor              | {On/Off} |
+| Git Branching        | {None/Per Stage/Per Milestone} |
+
+These settings apply to future /renn.plan-stage and /renn.run-stage runs.
+
+Quick commands:
+- /renn.set-profile <profile> — switch horsepower profile
+- /renn.plan-stage --research — force research
+- /renn.plan-stage --skip-research — skip research
+- /renn.plan-stage --skip-verify — skip plan review
+```
+
+</process>
+
+<success_criteria>
+- [ ] Current config read
+- [ ] User presented with 5 settings (horsepower + 3 workflow toggles + git branching)
+- [ ] Config updated with horsepower, checks, and git sections
+- [ ] Changes confirmed to user
+</success_criteria>

package/commands/renn.ship.md

@@ -0,0 +1,52 @@
+---
+name: renn.ship
+description: Ship your project to a deployment target
+allowed-tools:
+  - Read
+  - Bash
+  - Glob
+  - Grep
+  - Write
+  - Task
+  - AskUserQuestion
+---
+
+<objective>
+Prepare and ship your project to a deployment target (Vercel, npm, Railway, etc.).
+
+Reads brief.md for project context, detects stack for smart platform suggestions, and delegates to ship-project workflow.
+
+**Three phases:**
+1. ASK -- Present project summary, suggest platforms, user declares target
+2. RESEARCH & PLAN -- Investigate target requirements, generate checklist
+3. WALK CHECKLIST -- Execute auto items, present gates for human steps, track progress
+
+When `.renn/ship-plan.md` exists from a prior run, offers resume/restart/different-target.
+</objective>
+
+<execution_context>
+@~/.claude/renn/workflows/ship-project.md
+</execution_context>
+
+<context>
+$ARGUMENTS
+
+@.renn/pulse.md
+@.renn/brief.md
+</context>
+
+<process>
+**Follow ship-project.md workflow.**
+</process>
+
+<success_criteria>
+- [ ] Project summary presented from brief.md
+- [ ] Smart platform suggestions based on detected stack
+- [ ] User declared shipping target (or resumed existing plan)
+- [ ] Target persisted to .renn/ship-target.md for Phase 2
+- [ ] Scout research spawned for target platform
+- [ ] Deployment checklist saved to .renn/ship-plan.md with auto/gate classification
+- [ ] Checklist items executed (auto) or presented (gate) with progress tracking
+- [ ] Failed steps offer retry/skip/abort recovery
+- [ ] Ship workflow completes with summary of results
+</success_criteria>