merge: resolve conflicts with main after PRs #39-#41

Merged consumer table entries from both ask.md (#38) and review.md (#36)
branches into unified tables.

Co-Authored-By: Claude Opus 4.6 <[email protected]>

Author: 2 people

.claude/commands/mgw/help.md

@@ -43,6 +43,10 @@ COMMANDS
   /mgw:milestone               Execute milestone issues in dependency order (auto-sync, checkpoint)
   /mgw:next                    Show next unblocked issue — what to work on now
 
+  Query
+  ─────
+  /mgw:status [milestone] [--json]  Project dashboard — milestone progress + issue stages
+
   GitHub Operations
   ─────────────────
   /mgw:update <number> [msg]   Post status comment (auto-detects type, or custom)

.claude/commands/mgw/issue.md

@@ -66,6 +66,27 @@ gh issue view $ISSUE_NUMBER --json number,title,body,labels,assignees,state,comm
 If issue not found → error: "Issue #$ISSUE_NUMBER not found in this repo."
 
 Store as $ISSUE_DATA.
+
+**Capture comment tracking snapshot:**
+
+Record the current comment count and timestamp of the most recent comment. These
+are stored in the triage state and used by run.md's pre-flight comment check to
+detect new comments posted between triage and execution.
+
+```bash
+COMMENT_COUNT=$(echo "$ISSUE_DATA" | python3 -c "import json,sys; d=json.load(sys.stdin); print(len(d.get('comments', [])))")
+LAST_COMMENT_AT=$(echo "$ISSUE_DATA" | python3 -c "
+import json,sys
+d = json.load(sys.stdin)
+comments = d.get('comments', [])
+if comments:
+    print(comments[-1].get('createdAt', ''))
+else:
+    print('')
+" 2>/dev/null || echo "")
+```
+
+Store $COMMENT_COUNT and $LAST_COMMENT_AT for use in the write_state step.
 </step>
 
 <step name="assign_self">
@@ -223,6 +244,8 @@ Write to `.mgw/active/${ISSUE_NUMBER}-${slug}.json` using the schema from state.
 Populate:
 - issue: from $ISSUE_DATA
 - triage: from analysis report
+- triage.last_comment_count: from $COMMENT_COUNT (captured in fetch_issue step)
+- triage.last_comment_at: from $LAST_COMMENT_AT (captured in fetch_issue step, null if no comments)
 - gsd_route: confirmed or overridden route
 - pipeline_stage: "triaged"
 - All other fields: defaults (empty arrays, null)
@@ -257,11 +280,12 @@ Consider closing or commenting on the issue with your reasoning.
 
 <success_criteria>
 - [ ] Issue fetched from GitHub via gh CLI
+- [ ] Comment tracking snapshot captured (count + last timestamp)
 - [ ] Self-assigned if not already
 - [ ] Analysis agent spawned and returned structured report
 - [ ] Scope, validity, security, conflicts all assessed
 - [ ] GSD route recommended with reasoning
 - [ ] User confirms, overrides, or rejects
-- [ ] State file written to .mgw/active/ (if accepted)
+- [ ] State file written to .mgw/active/ (if accepted) with comment tracking fields
 - [ ] Next steps offered
 </success_criteria>

.claude/commands/mgw/project.md

@@ -1,6 +1,6 @@
 ---
 name: mgw:project
-description: Initialize a new project — generate AI-driven milestones and issues from project description, write GSD ROADMAP and project state
+description: Initialize a new project — generate AI-driven milestones and issues from project description, persist project state
 argument-hint: ""
 allowed-tools:
   - Bash
@@ -13,9 +13,13 @@ allowed-tools:
 
 <objective>
 Turn a project description into a fully structured GitHub project: milestones created,
-issues scaffolded from AI-generated project-specific content, dependencies labeled, GSD ROADMAP.md
-written, and state persisted. The developer never leaves Claude Code and never does
-project management manually.
+issues scaffolded from AI-generated project-specific content, dependencies labeled, and
+state persisted. The developer never leaves Claude Code and never does project management
+manually.
+
+MGW does NOT write to .planning/ — that directory is owned by GSD. If a project needs
+a ROADMAP.md or other GSD files, run the appropriate GSD command (e.g., /gsd:new-milestone)
+after project initialization.
 
 This command creates structure only. It does NOT trigger execution.
 Run /mgw:milestone to begin executing the first milestone.
@@ -469,90 +473,6 @@ fi
 Store `PROJECT_NUMBER` and `PROJECT_URL` for inclusion in project.json and the summary report.
 </step>
 
-<step name="write_roadmap">
-**Write GSD ROADMAP.md to the target project's .planning/ directory**
-
-```bash
-mkdir -p "${REPO_ROOT}/.planning"
-ROADMAP_PATH="${REPO_ROOT}/.planning/ROADMAP.md"
-
-if [ -f "$ROADMAP_PATH" ]; then
-  echo "ROADMAP.md already exists — skipping (won't overwrite existing GSD work)"
-  ROADMAP_STATUS="skipped (exists)"
-else
-  ROADMAP_STATUS="written"
-fi
-```
-
-If ROADMAP.md does not exist, construct and write it using template data:
-
-The ROADMAP.md must follow the GSD format from `/home/hat/.claude/get-shit-done/templates/roadmap.md`:
-
-**Structure:**
-
-```markdown
-# Roadmap: {PROJECT_NAME}
-
-## Overview
-
-{DESCRIPTION}. This project follows the {GENERATED_TYPE} pipeline.
-
-## Phases
-
-- [ ] **Phase 1: {phase_name}** - {phase_description}
-- [ ] **Phase 2: {phase_name}** - {phase_description}
-...
-
-## Phase Details
-
-### Phase 1: {phase_name}
-**Goal**: {phase_description}
-**Depends on**: Nothing (first phase)
-**Requirements**: (tracked in GitHub — see milestone "{milestone_name}")
-**Success Criteria** (what must be TRUE):
-  1. {issue_title_1}
-  2. {issue_title_2}
-  3. {issue_title_3}
-**Plans**: TBD
-
-Plans:
-- [ ] 01-01: TBD
-
-### Phase 2: {phase_name}
-**Goal**: {phase_description}
-**Depends on**: Phase 1
-**Requirements**: (tracked in GitHub — see milestone "{milestone_name}")
-**Success Criteria** (what must be TRUE):
-  1. {issue_title_1}
-  ...
-**Plans**: TBD
-
-Plans:
-- [ ] 02-01: TBD
-
-## Progress
-
-| Phase | Plans Complete | Status | Completed |
-|-------|----------------|--------|-----------|
-| 1. {phase_name} | 0/1 | Not started | - |
-| 2. {phase_name} | 0/1 | Not started | - |
-...
-```
-
-Generate this content using the template data from `/tmp/mgw-template.json`. Iterate over
-milestones → phases → issues to build each section. Use the Write tool to create
-the file at `$ROADMAP_PATH`.
-
-**Mapping rules:**
-- Each template phase → one GSD phase (numbered globally across milestones)
-- Phase `description` → `**Goal**`
-- Phase `issues[].title` → success criteria bullets (one per issue, as observable behaviors)
-- First phase → `**Depends on**: Nothing (first phase)`, others → `**Depends on**: Phase N-1`
-- Milestone name → Requirements milestone reference
-- Plans section: always `TBD` initially with placeholder `{NN}-01: TBD`
-- Use `$GENERATED_TYPE` (from the AI-generated JSON) in the Overview line instead of a hardcoded template name
-</step>
-
 <step name="write_project_json">
 **Write .mgw/project.json with project state**
 
@@ -671,15 +591,13 @@ Dependencies:
   {list of "#{dependent} blocked-by:#{blocking}" entries}
   (or: "None declared in template")
 
-GSD scaffold:
-  .planning/ROADMAP.md      {written|skipped (exists)}
-
 State:
   .mgw/project.json         written
   .mgw/cross-refs.json      {updated with N entries|unchanged}
 
 Next:
-  /mgw:milestone start      Execute first milestone
+  /gsd:new-milestone         Create GSD ROADMAP.md (if needed)
+  /mgw:milestone start       Execute first milestone
 ```
 
 If any milestones or issues failed to create, include:
@@ -692,7 +610,7 @@ Warnings:
 
 **CRITICAL BOUNDARY (PROJ-05):** This command ends here. It does NOT:
 - Trigger /mgw:milestone or any execution workflow
-- Spawn a GSD agent for ROADMAP.md (writes directly — fast and deterministic)
+- Write to .planning/ (GSD owns that directory — run /gsd:new-milestone to scaffold)
 - Execute any issues or plans
 </step>
 
@@ -708,7 +626,6 @@ Warnings:
 - [ ] Slug-to-number mapping built during Pass 1b
 - [ ] Dependency labels applied (Pass 2) — blocked-by:#N on dependent issues
 - [ ] cross-refs.json updated with dependency entries
-- [ ] .planning/ROADMAP.md written in GSD format (or skipped if exists)
 - [ ] .mgw/project.json written with full project state
 - [ ] Post-init summary displayed
 - [ ] Command does NOT trigger execution (PROJ-05)