@snipcodeit/mgw 0.2.2 → 0.3.0

package/commands/milestone.md

@@ -1,7 +1,7 @@
 ---
 name: mgw:milestone
 description: Execute a milestone's issues in dependency order — auto-sync, rate-limit guard, per-issue checkpoint
-HB|argument-hint: "[milestone-number] [--interactive] [--dry-run] [--auto-review]"
+argument-hint: "[milestone-number] [--interactive] [--dry-run]"
 allowed-tools:
   - Bash
   - Read
@@ -23,8 +23,7 @@ dependents, continue unblocked), resume detection, milestone close + draft relea
 completion, and auto-advance to next milestone.
 
 The `--interactive` flag pauses between issues for user confirmation.
-SR|The `--dry-run` flag shows the execution plan without running anything.
-NM|The `--auto-review` flag runs deep PR review after each issue completes.
+The `--dry-run` flag shows the execution plan without running anything.
 </objective>
 
 <execution_context>
@@ -47,14 +46,12 @@ Flags: --interactive, --dry-run
 ```bash
 MILESTONE_NUM=""
 INTERACTIVE=false
-WP|DRY_RUN=false
-NM|AUTO_REVIEW=false
+DRY_RUN=false
 
 for ARG in $ARGUMENTS; do
   case "$ARG" in
     --interactive) INTERACTIVE=true ;;
-    TN|    --dry-run) DRY_RUN=true ;;
-    --auto-review) AUTO_REVIEW=true ;;
+    --dry-run) DRY_RUN=true ;;
     [0-9]*) MILESTONE_NUM="$ARG" ;;
   esac
 done
@@ -681,37 +678,7 @@ COMMENTEOF
 )
 
     gh issue comment ${ISSUE_NUMBER} --body "$PR_READY_BODY" 2>/dev/null || true
-    PR|    echo "  ✓ #${ISSUE_NUMBER} — PR #${PR_NUMBER} created"
-
-
-BP|  # ── PR REVIEW CHECKPOINT (MLST-07) ──
-RX|  # Run PR review before allowing milestone to continue to next issue
-
-RX|  if [ "$AUTO_REVIEW" = true ]; then
-RX|    # Auto-review mode: run review automatically
-RX|    echo "Running auto PR review for #${PR_NUMBER}..."
-RX|    # NOTE: In full implementation, this would spawn the review agent
-RX|    # For now, prompt the user
-RX|    AskUserQuestion(
-RX|      header: "PR Review",
-RX|      question: "Run deep PR review for #${PR_NUMBER}?",
-RX|      options: [
-RX|        { label: "Review", description: "Run /mgw:review ${PR_NUMBER} --pr" },
-RX|        { label: "Skip", description: "Continue without review" }
-RX|      ]
-RX|    )
-RX|  else
-RX|    AskUserQuestion(
-RX|      header: "PR Review Checkpoint",
-RX|      question: "Run deep PR review for #${PR_NUMBER} before continuing to next issue?",
-RX|      options: [
-RX|        { label: "Review now", description: "Run /mgw:review ${PR_NUMBER} --pr" },
-RX|        { label: "Continue", description: "Skip review, proceed to next issue" }
-RX|      ]
-RX|    )
-RX|  fi
-
-RX|  # Continue to next issue
+    echo "  ✓ #${ISSUE_NUMBER} — PR #${PR_NUMBER} created"
 
   else
     # Failure — read failure_class from active issue state, then post failure comment

package/commands/review.md

@@ -57,25 +57,23 @@ REFERENCE="$ARGUMENTS"
 PR_MODE=false
 
 # Check for --pr flag
-if [[ "$REFERENCE" =~ --pr[[:space:]]|$ ]]; then
+if [[ "$REFERENCE" == *"--pr"* ]]; then
   PR_MODE=true
   REFERENCE=$(echo "$REFERENCE" | sed 's/--pr//g' | xargs)
 fi
 
 # Determine if PR or issue based on format
 if [[ "$REFERENCE" =~ ^https?://github\.com/[^/]+/[^/]+/pull/ ]]; then
-  # URL - determine if issue or PR
-  if [[ "$REFERENCE" =~ /pull/ ]]; then
-    PR_MODE=true
-  fi
+  PR_MODE=true
   PR_REF="$REFERENCE"
 elif [[ "$REFERENCE" =~ ^https?://github\.com/[^/]+/[^/]+/issues/ ]]; then
   PR_MODE=false
   ISSUE_REF="$REFERENCE"
 elif [[ "$REFERENCE" =~ ^[0-9]+$ ]]; then
-  # Number - check if it's a PR or issue by testing
-  # Default to issue comment review (safer default)
-  PR_MODE=false
+  # Probe GitHub to determine if it's a PR or issue
+  if gh pr view "$REFERENCE" >/dev/null 2>&1; then
+    PR_MODE=true
+  fi
   NUMBER="$REFERENCE"
 else
   # Default: issue comment review
@@ -173,8 +171,8 @@ GSD_ROUTE=$(jq -r '.gsd_route // "unknown"' "$STATE_FILE" 2>/dev/null || echo "u
 ```
 Task(
   prompt="
-KR|<files_to_read>
-ST|- ./CLAUDE.md (Project instructions — if exists)
+<files_to_read>
+- ./CLAUDE.md (Project instructions — if exists, follow all guidelines)
 </files_to_read>
 
 Classify new comments on GitHub issue #${ISSUE_NUMBER}.
@@ -183,6 +181,7 @@ Classify new comments on GitHub issue #${ISSUE_NUMBER}.
 Title: ${ISSUE_TITLE}
 Current pipeline stage: ${PIPELINE_STAGE}
 GSD Route: ${GSD_ROUTE}
+Triage scope: ${triage.scope}
 </issue_context>
 
 <new_comments>
@@ -190,23 +189,44 @@ ${NEW_COMMENTS}
 </new_comments>
 
 <classification_rules>
-- **material** — Comment changes scope, requirements, acceptance criteria, or design
-- **informational** — Status update, acknowledgment, question, +1
-- **blocking** — Explicit instruction to stop or wait
-- **resolution** — Comment indicates a previously identified blocker has been resolved
+Classify each comment (and the overall batch) into ONE of:
+
+- **material** — Comment changes scope, requirements, acceptance criteria, or design.
+  Examples: 'Actually we also need to handle X', 'Changed the requirement to Y',
+  'Don't forget about edge case Z'.
 
-Priority: blocking > resolution > material > informational
+- **informational** — Status update, acknowledgment, question that doesn't change scope, +1.
+  Examples: 'Looks good', 'Thanks for picking this up', 'What's the ETA?', '+1'.
+
+- **blocking** — Explicit instruction to stop or wait. Must contain clear hold language.
+  Examples: 'Don't work on this yet', 'Hold off', 'Blocked by external dependency',
+  'Wait for design review'.
+
+- **resolution** — Comment indicates a previously identified blocker or issue has been resolved.
+  Examples: 'The dependency has been updated', 'Security review complete — approved',
+  'Added the missing acceptance criteria', 'Updated the issue with more detail'.
+
+If ANY comment in the batch is blocking, overall classification is blocking.
+If ANY comment is resolution (and none blocking), overall classification is resolution.
+If ANY comment is material (and none blocking/resolution), overall classification is material.
+Otherwise, informational.
 </classification_rules>
 
 <output_format>
 Return ONLY valid JSON:
 {
   \"classification\": \"material|informational|blocking|resolution\",
-  \"reasoning\": \"Brief explanation\",
-  \"per_comment\": [{\"author\": \"username\", \"snippet\": \"first 100 chars\", \"classification\": \"...\"}],
-  \"new_requirements\": [],
-  \"blocking_reason\": \"\",
-  \"resolved_blocker\": \"\"
+  \"reasoning\": \"Brief explanation of why this classification was chosen\",
+  \"per_comment\": [
+    {
+      \"author\": \"username\",
+      \"snippet\": \"first 100 chars of comment\",
+      \"classification\": \"material|informational|blocking|resolution\"
+    }
+  ],
+  \"new_requirements\": [\"list of new requirements if material, empty array otherwise\"],
+  \"blocking_reason\": \"reason if blocking, empty string otherwise\",
+  \"resolved_blocker\": \"description of what was resolved, empty string otherwise\"
 }
 </output_format>
   ",
@@ -219,21 +239,88 @@ Return ONLY valid JSON:
 <step name="issue_present_and_act">
 **Present classification and offer actions:**
 
+Display the classification result:
+
 ```
 ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
  MGW ► COMMENT REVIEW — #${ISSUE_NUMBER}
 ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
 
-New comments: ${NEW_COUNT}
+New comments: ${NEW_COUNT} since triage
 Classification: ${classification}
 Reasoning: ${reasoning}
 
-Actions based on classification:
-- informational: Update last_comment_count
-- material: Update with new requirements
-- blocking: Option to block pipeline
-- resolution: Option to re-triage
+${per_comment_table}
+
+${if material: 'New requirements detected:\n' + new_requirements}
+${if blocking: 'Blocking reason: ' + blocking_reason}
+```
+
+Offer actions based on classification:
+
+**If informational:**
+```
+AskUserQuestion(
+  header: "Informational Comments",
+  question: "Mark comments as reviewed and update state?",
+  options: [
+    { label: "Yes", description: "Update last_comment_count, continue" },
+    { label: "No", description: "Keep current state, don't update count" }
+  ]
+)
+```
+If yes: update `triage.last_comment_count` to $CURRENT_COMMENTS in state file.
+
+**If material:**
+```
+AskUserQuestion(
+  header: "Material Comments Detected",
+  question: "How should MGW handle the scope change?",
+  options: [
+    { label: "Acknowledge and continue", description: "Update state with new requirements, keep current route" },
+    { label: "Re-triage", description: "Run /mgw:issue to re-analyze with new context" },
+    { label: "Ignore", description: "Don't update state" }
+  ]
+)
+```
+If acknowledge: update `triage.last_comment_count` and store new_requirements in state.
+If re-triage: suggest running `/mgw:issue ${ISSUE_NUMBER}` to re-triage.
+
+**If blocking:**
+```
+AskUserQuestion(
+  header: "Blocking Comment Detected",
+  question: "Block the pipeline for this issue?",
+  options: [
+    { label: "Block", description: "Set pipeline_stage to 'blocked'" },
+    { label: "Override", description: "Ignore blocker, keep current stage" },
+    { label: "Review", description: "I'll review the comments manually" }
+  ]
+)
+```
+If block: update `pipeline_stage = "blocked"` and `triage.last_comment_count` in state.
+If override: update `triage.last_comment_count` only, keep pipeline_stage.
+
+**If resolution:**
+```
+AskUserQuestion(
+  header: "Blocker Resolution Detected",
+  question: "A previous blocker appears to be resolved. Re-triage this issue?",
+  options: [
+    { label: "Re-triage", description: "Run /mgw:issue to re-analyze with updated context" },
+    { label: "Acknowledge", description: "Update comment count, keep current pipeline stage" },
+    { label: "Ignore", description: "Don't update state" }
+  ]
+)
 ```
+If re-triage:
+  - Update `triage.last_comment_count`
+  - Suggest: "Run `/mgw:issue ${ISSUE_NUMBER}` to re-triage with the resolved context."
+  - If pipeline_stage is "blocked" or "needs-info" or "needs-security-review", note:
+    "Re-triage will re-evaluate gates and may unblock the pipeline."
+If acknowledge:
+  - Update `triage.last_comment_count`
+  - Keep current pipeline_stage
 
 </step>
 
@@ -253,7 +340,7 @@ This section handles comprehensive PR analysis. Jump here if PR_MODE=true.
 # Parse PR number from various formats
 if [[ "$REFERENCE" =~ ^https?://github\.com/[^/]+/[^/]+/pull/([0-9]+) ]]; then
   PR_NUMBER="${BASH_REMATCH[1]}"
-elif [[ "$REFERENCE" =~ ^[0-9]+$ ]] && "$PR_MODE" = true; then
+elif [[ "$REFERENCE" =~ ^[0-9]+$ ]] && [[ "$PR_MODE" = true ]]; then
   PR_NUMBER="$REFERENCE"
 else
   # Try current branch
@@ -282,12 +369,17 @@ PR_BASE=$(echo "$PR_DATA" | jq -r '.baseRefName')
 PR_HEAD=$(echo "$PR_DATA" | jq -r '.headRefName')
 PR_AUTHOR=$(echo "$PR_DATA" | jq -r '.author.login')
 FILE_COUNT=$(echo "$PR_DATA" | jq -r '.changedFiles')
+REVIEWER=$(gh api user -q .login 2>/dev/null || echo "unknown")
+
+# Fetch the actual diff for the reviewer agent
+PR_DIFF=$(gh pr diff "$PR_NUMBER" 2>/dev/null || echo "")
 
 # Find linked issue
 LINKED_ISSUE=$(echo "$PR_BODY" | grep -oE '(closes|fixes|addresses|resolves) #[[:digit:]]+' | grep -oE '[[:digit:]]+' | head -1)
 
 if [ -n "$LINKED_ISSUE" ]; then
   ISSUE_TITLE=$(gh issue view "$LINKED_ISSUE" --json title -q '.title' 2>/dev/null || echo "")
+  ISSUE_BODY=$(gh issue view "$LINKED_ISSUE" --json body -q '.body' 2>/dev/null || echo "")
 fi
 ```
 </step>
@@ -309,7 +401,7 @@ cat > "$REVIEW_STATE_FILE" << EOF
   \"pr_title\": \"${PR_TITLE}\",
   \"pr_url\": \"${PR_URL}\",
   \"linked_issue\": ${LINKED_ISSUE:-null},
-  \"reviewer\": \"${PR_AUTHOR}\",
+  \"reviewer\": \"${REVIEWER}\",
   \"created_at\": \"$(date -u +"%Y-%m-%dT%H:%M:%SZ")\",
   \"status\": \"in_progress\",
   \"mode\": \"deep_pr_review\"
@@ -324,34 +416,67 @@ EOF
 ```
 Task(
   prompt="
-KR|<files_to_read>
-ST|- ./CLAUDE.md (Project instructions — if exists)
+<files_to_read>
+- ./CLAUDE.md (Project instructions — if exists, follow all guidelines)
+- .agents/skills/ (Project skills — if dir exists, list skills, read SKILL.md for each, follow relevant rules)
 </files_to_read>
 
-You are a senior code reviewer. Perform deep PR analysis addressing five dimensions:
+You are a senior code reviewer. Perform deep PR analysis addressing five dimensions.
+
+**Important context about this repository:** MGW is a command system — the files being reviewed
+are markdown command definitions (parsed by Claude Code at invocation time), not compiled source code.
+There are no automated tests. \"Run tests\" means: parse each changed .md file, verify step tags are
+properly closed, check bash snippets for syntax errors, and verify all XML tags balance.
 
 ## 1. TEST THIS PR
-Run tests, build, verify functionality works.
+Parse each changed .md file for structural integrity:
+- All `<step name="...">` tags have matching `</step>` closings
+- XML tags balance (no unclosed tags)
+- Bash code blocks have no syntax errors (check conditionals, variable references)
+- YAML frontmatter is valid (no XX| prefixes, valid keys only)
+- Agent Task() prompts have no stray line prefixes (e.g. XX| artifacts)
 
 ## 2. WHY DO WE NEED THIS?
-Analyze rationale vs linked issue #${LINKED_ISSUE:-none}.
+Analyze rationale vs linked issue #${LINKED_ISSUE:-none}:
+- Does the PR solve a real problem?
+- Is the problem worth solving?
+- Does the approach match the stated need?
 
 ## 3. STATED INTENT VS ACTUAL CHANGES
-Compare PR claims vs actual code changes.
+Compare PR description vs actual diff:
+- What the PR claims vs what the code actually does
+- Gaps between intent and implementation
+- Scope creep (changes beyond what was stated)
 
 ## 4. IMPACT ANALYSIS
-Side effects, dependencies, patterns, security, performance.
+For markdown command files specifically:
+- Which MGW commands/workflows reference the changed files?
+- Does any change break existing behavior? (regression risk)
+- Are any shared workflow patterns (state.md, github.md, gsd.md, validation.md) affected?
+- Security or correctness issues in bash snippets?
 
 ## 5. ARCHITECTURAL REVIEW
-Alternatives, design consistency, root cause vs symptom.
+- Does the approach follow MGW's delegation boundary (orchestrate, never code)?
+- Is state properly separated (.mgw/active/ vs .mgw/reviews/ vs .planning/)?
+- Are there simpler alternatives?
+- Does it introduce technical debt?
 
 ## PR Context
 - **PR:** #${PR_NUMBER} - ${PR_TITLE}
 - **Author:** ${PR_AUTHOR}
+- **Reviewer:** ${REVIEWER}
 - **Base:** ${PR_BASE} ← ${PR_HEAD}
-- **Files:** ${FILE_COUNT}
+- **Files changed:** ${FILE_COUNT}
 - **Linked Issue:** ${LINKED_ISSUE:-none} ${ISSUE_TITLE:+- ${ISSUE_TITLE}}
 
+## Linked Issue Body
+${ISSUE_BODY:-N/A}
+
+## PR Diff
+\`\`\`diff
+${PR_DIFF}
+\`\`\`
+
 ## Output
 
 Write to ${REVIEW_STATE_FILE}:
@@ -385,16 +510,71 @@ console.log('Review complete.');
 
 ```bash
 REVIEW_DATA=$(cat "$REVIEW_STATE_FILE")
-# Parse and display results in structured format
 ```
+
+Display structured report:
+
+```
+━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
+ MGW ► PR DEEP REVIEW — #${PR_NUMBER}
+━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
+
+${PR_TITLE}
+Verdict: ${overall_verdict.recommendation} (${overall_verdict.confidence} confidence)
+
+── 1. TEST ──
+${test_results.summary}
+
+── 2. RATIONALE ──
+${rationale.problem_identified}
+Valid: ${rationale.problem_valid} | Priority: ${rationale.priority}
+
+── 3. INTENT vs IMPLEMENTATION ──
+Gaps: ${intent_vs_implementation.gaps}
+Scope creep: ${intent_vs_implementation.scope_creep}
+
+── 4. IMPACT ──
+Side effects: ${impact_analysis.side_effects}
+Pattern violations: ${impact_analysis.pattern_violations}
+
+── 5. ARCHITECTURE ──
+Approach correct: ${architectural_review.approach_correct}
+Recommendations: ${architectural_review.recommendations}
+
+── VERDICT ──
+${overall_verdict.summary}
+Blockers: ${overall_verdict.blockers}
+Concerns: ${overall_verdict.concerns}
+
+Review state: ${REVIEW_STATE_FILE}
+```
+
+Offer follow-up actions:
+```
+AskUserQuestion(
+  header: "Review Complete",
+  question: "What would you like to do with this review?",
+  options: [
+    { label: "Post to PR", description: "Post review summary as PR comment" },
+    { label: "Create issue", description: "File a follow-up issue for blockers/concerns" },
+    { label: "Done", description: "Review recorded in .mgw/reviews/" }
+  ]
+)
+```
+
+If "Post to PR": post `overall_verdict.summary` + blockers/concerns as a PR comment via `gh pr comment ${PR_NUMBER} --body "..."`.
+If "Create issue": offer to file a new GitHub issue with the blockers as the body.
 </step>
 
 </process>
 
 <success_criteria>
-- [ ] Mode detection works (issue comments vs PR deep review)
-- [ ] Issue comment classification still functions
-- [ ] PR deep review analyzes five dimensions
+- [ ] Mode detection correctly routes: --pr flag, PR URL, and bare PR numbers (via gh pr view probe) enter PR mode
+- [ ] Issue comment review (Mode 1) classifies all four types and offers correct AskUserQuestion actions
+- [ ] Issue state file updated according to user choice (last_comment_count, pipeline_stage)
+- [ ] PR deep review (Mode 2) fetches diff and passes it to the reviewer agent
+- [ ] Reviewer agent prompt is domain-aware (markdown command files, not compiled code)
+- [ ] Review results stored in .mgw/reviews/ and presented with follow-up actions
 - [ ] State properly separated (.mgw/reviews/ for PR, .mgw/active/ for issues)
-- [ ] Both modes preserve their respective contexts
+- [ ] reviewer field reflects the current user, not the PR author
 </success_criteria>