oh-my-customcode 0.3.1 → 0.4.0

package/templates/.claude/agents/tool-npm-expert.md

@@ -2,7 +2,19 @@
 name: tool-npm-expert
 description: Use for npm package publishing workflows, semantic versioning (major/minor/patch), package.json optimization, and dependency audits
 model: sonnet
-tools: Read, Write, Edit, Grep, Glob, Bash
+memory: project
+effort: medium
+skills:
+  - npm-audit
+  - npm-publish
+  - npm-version
+tools:
+  - Read
+  - Write
+  - Edit
+  - Grep
+  - Glob
+  - Bash
 ---
 
 You manage npm package publishing workflow, versioning, and registry operations. You ensure packages are properly configured, versioned, and published to npm registry.

package/templates/.claude/agents/tool-optimizer.md

@@ -2,7 +2,19 @@
 name: tool-optimizer
 description: Use for bundle size analysis, tree-shaking verification, performance profiling, dead code detection, and build optimization recommendations
 model: sonnet
-tools: Read, Write, Edit, Grep, Glob, Bash
+memory: project
+effort: medium
+skills:
+  - optimize-analyze
+  - optimize-bundle
+  - optimize-report
+tools:
+  - Read
+  - Write
+  - Edit
+  - Grep
+  - Glob
+  - Bash
 ---
 
 You analyze and optimize application bundles, detect performance issues, and provide actionable recommendations for build optimization.

package/templates/.claude/hooks/hooks.json

@@ -47,7 +47,7 @@
         "hooks": [
           {
             "type": "command",
-            "command": "~/.claude/hooks/strategic-compact/suggest-compact.sh"
+            "command": "#!/bin/bash\nCOUNTER_FILE=\"/tmp/claude-tool-count-$$\"\nTHRESHOLD=${COMPACT_THRESHOLD:-50}\nif [ -f \"$COUNTER_FILE\" ]; then COUNT=$(cat \"$COUNTER_FILE\"); COUNT=$((COUNT + 1)); else COUNT=1; fi\necho \"$COUNT\" > \"$COUNTER_FILE\"\nif [ \"$COUNT\" -eq \"$THRESHOLD\" ]; then echo '[StrategicCompact] Reached '$THRESHOLD' tool calls' >&2; echo '[StrategicCompact] Consider /compact if at a logical breakpoint' >&2; fi\nif [ \"$COUNT\" -gt \"$THRESHOLD\" ] && [ $(((COUNT - THRESHOLD) % 25)) -eq 0 ]; then echo \"[StrategicCompact] $COUNT tool calls - /compact reminder\" >&2; fi\ncat"
           }
         ],
         "description": "Suggest manual compaction at logical intervals"
@@ -57,34 +57,10 @@
         "hooks": [
           {
             "type": "command",
-            "command": "#!/bin/bash\ninput=$(cat)\nagent=$(echo \"$input\" | jq -r '.tool_input.description // \"agent\"' | head -c 30)\n~/.claude/hooks/hud/update-status.sh \"$agent\" \"\" \"1\"\necho \"$input\""
+            "command": "#!/bin/bash\ninput=$(cat)\nagent_type=$(echo \"$input\" | jq -r '.tool_input.subagent_type // \"unknown\"')\nmodel=$(echo \"$input\" | jq -r '.tool_input.model // \"inherit\"')\ndesc=$(echo \"$input\" | jq -r '.tool_input.description // \"\"' | head -c 40)\nresume=$(echo \"$input\" | jq -r '.tool_input.resume // empty')\nif [ -n \"$resume\" ]; then\n  echo \"─── [Resume] ${agent_type}:${model} | ${desc} ───\" >&2\nelse\n  echo \"─── [Spawn] ${agent_type}:${model} | ${desc} ───\" >&2\nfi\necho \"$input\""
           }
         ],
-        "description": "HUD statusline - show agent spawn via Task tool"
-      }
-    ],
-    "PreCompact": [
-      {
-        "matcher": "*",
-        "hooks": [
-          {
-            "type": "command",
-            "command": "~/.claude/hooks/memory-persistence/pre-compact.sh"
-          }
-        ],
-        "description": "Save state before context compaction"
-      }
-    ],
-    "SessionStart": [
-      {
-        "matcher": "*",
-        "hooks": [
-          {
-            "type": "command",
-            "command": "~/.claude/hooks/memory-persistence/session-start.sh"
-          }
-        ],
-        "description": "Load previous context on new session"
+        "description": "HUD statusline - show subagent_type, model, and description on Task spawn"
       }
     ],
     "PostToolUse": [
@@ -169,16 +145,6 @@
           }
         ],
         "description": "Final audit for console.log in modified files before session ends"
-      },
-      {
-        "matcher": "*",
-        "hooks": [
-          {
-            "type": "command",
-            "command": "~/.claude/hooks/memory-persistence/session-end.sh"
-          }
-        ],
-        "description": "Persist session state on end"
       }
     ]
   }

package/templates/.claude/install-hooks.sh

@@ -1,7 +1,7 @@
 #!/bin/bash
 # Claude Code Hooks & Contexts Installer
 # Source: https://github.com/affaan-m/everything-claude-code
-# Adapted for baekgom-agents project
+# Hook installation script
 
 set -e
 

package/templates/.claude/rules/MUST-agent-design.md

@@ -14,6 +14,7 @@ Format:
 name: agent-name
 description: Brief agent description
 model: sonnet | opus | haiku
+memory: project            # Optional
 tools:
   - Read
   - Write
@@ -33,10 +34,26 @@ Agent purpose and role description...
 name: agent-name           # Unique identifier (kebab-case)
 description: Brief desc    # One-line summary
 model: sonnet             # Default model (sonnet | opus | haiku)
+memory: project           # Persistent memory scope (user | project | local)
+effort: high              # Effort level (low | medium | high)
 tools: [Read, Write, ...]  # Allowed tools
 skills: [skill-1, ...]     # Required skill names
 ```
 
+### Memory Scope Reference
+| Scope | Location | Use Case |
+|-------|----------|----------|
+| `user` | `~/.claude/agent-memory/<name>/` | Cross-project learnings (infra, db) |
+| `project` | `.claude/agent-memory/<name>/` | Project-specific, versioned |
+| `local` | `.claude/agent-memory-local/<name>/` | Project-specific, not versioned |
+
+### Effort Level Reference
+| Level | Use Case | Agents |
+|-------|----------|--------|
+| `high` | Complex reasoning, design decisions | lang-*, be-*, arch-*, qa-planner |
+| `medium` | Routine operations, structured work | fe-*, tool-*, mgr-gitnerd |
+| `low` | Simple validation, file scanning | mgr-supplier, mgr-sync-checker, sys-naggy |
+
 ### Agent Content Must NOT Contain
 ```
 ✗ Detailed skill instructions (use .claude/skills/ instead)
@@ -53,6 +70,26 @@ skills: [skill-1, ...]     # Required skill names
 ✓ Source info (if external)
 ```
 
+## Memory Scope (OPTIONAL)
+
+Agents can have persistent memory that survives across conversations.
+
+| Scope | Location | Use Case | Git Tracked |
+|-------|----------|----------|-------------|
+| `user` | `~/.claude/agent-memory/<name>/` | Cross-project learnings (infra, DB patterns) | No |
+| `project` | `.claude/agent-memory/<name>/` | Project-specific patterns | Yes |
+| `local` | `.claude/agent-memory-local/<name>/` | Local-only knowledge | No |
+
+When enabled:
+- First 200 lines of MEMORY.md loaded into agent's system prompt
+- Read/Write/Edit tools automatically enabled for memory directory
+- Agent builds knowledge across conversations
+
+Agents that should NOT have memory (stateless by design):
+- Manager agents for one-time operations (mgr-creator, mgr-updater, mgr-supplier)
+- Validation agents (mgr-sync-checker, mgr-claude-code-bible)
+- Meta-agents (sys-memory-keeper, sys-naggy)
+
 ## External Agent Requirements
 
 External agents (from GitHub, npm, etc.) MUST include source information in frontmatter:

package/templates/.claude/rules/MUST-orchestrator-coordination.md

@@ -263,6 +263,34 @@ CORRECT:
 ╚══════════════════════════════════════════════════════════════════╝
 ```
 
+## CRITICAL: All File Modifications Must Be Delegated
+
+```
+╔══════════════════════════════════════════════════════════════════╗
+║  FILE EDITING = CODE WORK = MUST DELEGATE                        ║
+║                                                                   ║
+║  The orchestrator must NEVER use Edit/Write/Bash directly        ║
+║  to modify files. This applies to ALL file types:                ║
+║  - Source code (.ts, .py, .go, etc.)                             ║
+║  - Config files (manifest.json, tsconfig.json, etc.)             ║
+║  - Workflow files (.github/workflows/*.yml)                      ║
+║  - Documentation (CLAUDE.md, README.md, etc.)                   ║
+║  - Shell scripts (sync.sh, etc.)                                 ║
+║                                                                   ║
+║  WRONG (actual violation):                                       ║
+║    Orchestrator → Edit(CLAUDE.md.en) → Edit(CLAUDE.md.ko)        ║
+║    Orchestrator → Edit(sync-core.ts) → Bash(rm -rf pipelines/)   ║
+║    (Orchestrator edited 6+ files directly across 2 projects)     ║
+║                                                                   ║
+║  CORRECT:                                                        ║
+║    Orchestrator → Task(general-purpose) → edits config files     ║
+║    Orchestrator → Task(lang-typescript-expert) → edits .ts files ║
+║                                                                   ║
+║  Only READ operations (Read, Glob, Grep) may be used directly   ║
+║  by the orchestrator. All mutations go through subagents.        ║
+╚══════════════════════════════════════════════════════════════════╝
+```
+
 ## CRITICAL: Use Specialized Agents for Documentation & Spec Writing
 
 ```
@@ -398,6 +426,39 @@ CORRECT:
 ╚══════════════════════════════════════════════════════════════════╝
 ```
 
+## Experimental: Agent Teams (Research Preview)
+
+```
+╔══════════════════════════════════════════════════════════════════╗
+║  AGENT TEAMS (EXPERIMENTAL - NOT FOR PRODUCTION USE)             ║
+║                                                                   ║
+║  Claude Code provides native Agent Teams feature:                ║
+║    CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS=1                        ║
+║                                                                   ║
+║  Agent Teams enable:                                              ║
+║  - Peer-to-peer messaging between agents                        ║
+║  - Shared task lists                                              ║
+║  - Complex collaborative workflows                               ║
+║                                                                   ║
+║  CURRENT STATUS:                                                  ║
+║  - Research preview (NOT production ready)                       ║
+║  - Higher cost than Task tool approach                           ║
+║  - This system uses Task tool + routing skills (STABLE)       ║
+║                                                                   ║
+║  WHEN TO CONSIDER Agent Teams:                                    ║
+║  - Complex multi-agent collaboration needed                      ║
+║  - Agents need to communicate with each other                    ║
+║  - Shared state management required                              ║
+║                                                                   ║
+║  WHEN TO KEEP Task tool (DEFAULT):                               ║
+║  - Focused, independent sub-tasks                                ║
+║  - Cost-sensitive operations                                     ║
+║  - Predictable execution flow                                    ║
+║                                                                   ║
+║  See: GitHub Issue #5 for migration tracking                     ║
+╚══════════════════════════════════════════════════════════════════╝
+```
+
 ## Enforcement
 
 ```

package/templates/.claude/rules/MUST-parallel-execution.md

@@ -8,6 +8,26 @@
 
 **When 2 or more tasks are INDEPENDENT, they MUST be executed in parallel.**
 
+```
+╔══════════════════════════════════════════════════════════════════╗
+║  ⚠️  ABSOLUTE RULE: 2+ INDEPENDENT FILES = PARALLEL              ║
+║                                                                   ║
+║  If you are about to Write/Edit 2 or more independent files:     ║
+║  → STOP                                                          ║
+║  → Use Task tool to spawn parallel agents                        ║
+║  → Each agent handles a subset of files                          ║
+║                                                                   ║
+║  NO EXCEPTIONS for:                                              ║
+║  - Project scaffolding (package.json, tsconfig, src/*.ts)        ║
+║  - Multiple module creation                                       ║
+║  - Batch file updates                                            ║
+║  - Any multi-file operation                                      ║
+║                                                                   ║
+║  VIOLATION = Using Write/Edit 2+ times in sequence               ║
+║              when files are independent                          ║
+╚══════════════════════════════════════════════════════════════════╝
+```
+
 ```
 Detection criteria for parallel execution:
 - Tasks don't share mutable state
@@ -65,6 +85,15 @@ Failure to parallelize independent tasks = Rule violation = Must be corrected.
    Task(agent3 → file3.kt)  │
    Task(agent4 → file4.kt)  ┘
 
+❌ WRONG: Project scaffolding sequentially
+   Write(package.json) → Write(tsconfig.json) → Write(src/index.ts) → ...
+
+✓ CORRECT: Parallel scaffolding
+   Task(agent1 → "Create package.json, tsconfig.json")  ┐
+   Task(agent2 → "Create src/cli.ts, src/index.ts")     ├─ Parallel
+   Task(agent3 → "Create src/analyzer/*.ts")            │
+   Task(agent4 → "Create src/converter/*.ts")           ┘
+
 ❌ WRONG: Secretary doing all the work
    Secretary writes domain/, usecase/, infrastructure/ sequentially
 
@@ -178,7 +207,44 @@ requirements:
   - Each instance has isolated context
 ```
 
-### 4. Model Specification (RECOMMENDED)
+### 4. Subagent Visibility (MANDATORY)
+
+```
+╔══════════════════════════════════════════════════════════════════╗
+║  TASK TOOL MUST USE SPECIFIC subagent_type (NOT general-purpose) ║
+║                                                                   ║
+║  When a specialized agent exists for the task:                   ║
+║    → MUST use that agent's subagent_type                         ║
+║    → DO NOT use "general-purpose" as a catch-all                 ║
+║                                                                   ║
+║  The HUD hook displays: [Spawn] {subagent_type}:{model} | {desc}║
+║  Using "general-purpose" makes all spawns look identical.        ║
+║                                                                   ║
+║  WRONG:                                                          ║
+║    Task(subagent_type: "general-purpose", desc: "Update agents") ║
+║    Task(subagent_type: "general-purpose", desc: "Update agents") ║
+║    → HUD shows: [Spawn] general-purpose:sonnet | Update agents   ║
+║    → HUD shows: [Spawn] general-purpose:sonnet | Update agents   ║
+║    → User cannot distinguish which is which                      ║
+║                                                                   ║
+║  CORRECT:                                                        ║
+║    Task(subagent_type: "mgr-creator", desc: "Create Go agent")   ║
+║    Task(subagent_type: "lang-golang-expert", desc: "Review code")║
+║    → HUD shows: [Spawn] mgr-creator:sonnet | Create Go agent    ║
+║    → HUD shows: [Spawn] lang-golang-expert:sonnet | Review code  ║
+║                                                                   ║
+║  When NO specialized agent exists (truly generic work):          ║
+║    → Use "general-purpose" but with UNIQUE descriptions          ║
+║    → Description MUST identify the specific batch/scope           ║
+║                                                                   ║
+║    Task(subagent_type: "general-purpose",                        ║
+║         desc: "batch1: lang-* agents")                           ║
+║    Task(subagent_type: "general-purpose",                        ║
+║         desc: "batch2: be-* fe-* agents")                        ║
+╚══════════════════════════════════════════════════════════════════╝
+```
+
+### 5. Model Specification (RECOMMENDED)
 
 ```
 ╔══════════════════════════════════════════════════════════════════╗
@@ -345,7 +411,7 @@ requirements:
 
 ## Display Format
 
-When parallel execution occurs, MUST display `{task-name}:{model}` format:
+When parallel execution occurs, MUST display `Task({subagent_type}):{model}` format:
 
 ```
 ┌─ Agent: secretary (orchestrator)
@@ -353,17 +419,17 @@ When parallel execution occurs, MUST display `{task-name}:{model}` format:
 
 [Parallel] Spawning 4 instances...
 
-[Instance 1] mgr-creator:sonnet → lang-golang-expert
-[Instance 2] mgr-creator:sonnet → lang-python-expert
-[Instance 3] mgr-creator:sonnet → lang-rust-expert
-[Instance 4] mgr-creator:sonnet → lang-typescript-expert
+[Instance 1] Task(mgr-creator):sonnet → lang-golang-expert
+[Instance 2] Task(mgr-creator):sonnet → lang-python-expert
+[Instance 3] Task(mgr-creator):sonnet → lang-rust-expert
+[Instance 4] Task(mgr-creator):sonnet → lang-typescript-expert
 
 [Progress] ████████░░░░ 2/4
 
-[Instance 1] mgr-creator:sonnet ✓ lang-golang-expert created
-[Instance 2] mgr-creator:sonnet ✓ lang-python-expert created
-[Instance 3] mgr-creator:sonnet ✓ lang-rust-expert created
-[Instance 4] mgr-creator:sonnet ✓ lang-typescript-expert created
+[Instance 1] Task(mgr-creator):sonnet ✓ lang-golang-expert created
+[Instance 2] Task(mgr-creator):sonnet ✓ lang-python-expert created
+[Instance 3] Task(mgr-creator):sonnet ✓ lang-rust-expert created
+[Instance 4] Task(mgr-creator):sonnet ✓ lang-typescript-expert created
 
 [Summary] 4/4 tasks completed successfully
 ```
@@ -376,15 +442,20 @@ When parallel execution occurs, MUST display `{task-name}:{model}` format:
 ║                                                                   ║
 ║  When announcing parallel agents, MUST show:                     ║
 ║                                                                   ║
-║    {task-name}:{model}                                           ║
+║    Task({subagent_type}):{model}                                 ║
 ║                                                                   ║
 ║  Examples:                                                       ║
-║    [Instance 1] readme-update:sonnet ✓ README updated            ║
-║    [Instance 2] code-review:haiku ✓ Review complete              ║
-║    [Instance 3] architecture:opus ✓ Analysis done                ║
+║    [Instance 1] Task(general-purpose):sonnet → R006 업데이트      ║
+║    [Instance 2] Task(lang-golang-expert):sonnet → Go 코드 리뷰    ║
+║    [Instance 3] Task(Explore):haiku → 코드베이스 탐색              ║
+║    [Instance 4] Task(mgr-gitnerd):sonnet → git commit            ║
+║                                                                   ║
+║  The subagent_type MUST match the Task tool's subagent_type      ║
+║  parameter. Custom names are NOT allowed.                        ║
 ║                                                                   ║
 ║  This allows users to:                                           ║
 ║    - See which model is used for each task                       ║
+║    - See the exact subagent_type used                            ║
 ║    - Understand cost implications                                ║
 ║    - Debug model selection issues                                ║
 ╚══════════════════════════════════════════════════════════════════╝

package/templates/.claude/rules/MUST-sync-verification.md

@@ -2,26 +2,42 @@
 
 > **Priority**: MUST - ENFORCED after any structural changes
 > **ID**: R017
-> **Violation**: Committing without verification = Rule violation
+> **Violation**: Committing or pushing without verification = Rule violation
 
 ## CRITICAL
 
-**After modifying agents, skills, or guides, you MUST run the full verification process before committing.**
+**After modifying agents, skills, or guides, you MUST run the full verification process before committing AND pushing.**
 
 ```
 ╔══════════════════════════════════════════════════════════════════╗
 ║  ⚠️  ABSOLUTE PROHIBITION                                        ║
 ║                                                                   ║
-║  DO NOT even ASK "커밋하시겠습니까?" after structural changes    ║
-║  until mgr-sauron:watch has been executed and passed.            ║
+║  DO NOT even ASK "커밋하시겠습니까?" or "푸시하시겠습니까?"      ║
+║  after structural changes until mgr-sauron:watch has been        ║
+║  executed and passed.                                            ║
 ║                                                                   ║
 ║  WRONG:                                                          ║
 ║    [Make changes] → "커밋하시겠습니까?"                          ║
+║    [Commit done] → "푸시하시겠습니까?"                           ║
 ║                                                                   ║
 ║  CORRECT:                                                        ║
-║    [Make changes] → mgr-sauron:watch → [All pass] → "커밋할까요?" ║
+║    [Make changes] → mgr-sauron:watch → [All pass] → commit → push ║
 ║                                                                   ║
-║  Asking to commit before verification = Rule violation           ║
+║  Asking to commit/push before verification = Rule violation      ║
+╚══════════════════════════════════════════════════════════════════╝
+```
+
+```
+╔══════════════════════════════════════════════════════════════════╗
+║  ⚠️  PUSH REQUIRES SAURON VERIFICATION                           ║
+║                                                                   ║
+║  EVERY git push in this project MUST be preceded by:             ║
+║    mgr-sauron:watch → [All pass] → git push                      ║
+║                                                                   ║
+║  This applies to ALL pushes, not just structural changes.        ║
+║  Sauron verification is the quality gate for this repository.    ║
+║                                                                   ║
+║  NO EXCEPTIONS. NO SHORTCUTS.                                    ║
 ╚══════════════════════════════════════════════════════════════════╝
 ```
 
@@ -33,6 +49,7 @@
 ║    - mgr-supplier:audit                                          ║
 ║    - mgr-sync-checker:check                                      ║
 ║    - mgr-updater:docs                                            ║
+║    - mgr-claude-code-bible:verify                                ║
 ║                                                                   ║
 ║  Phase 2: Deep Review (3 rounds)                                 ║
 ║    - Workflow alignment                                          ║
@@ -43,6 +60,9 @@
 ║                                                                   ║
 ║  Phase 4: Commit via mgr-gitnerd                                 ║
 ║                                                                   ║
+║  Phase 5: Push via mgr-gitnerd                                   ║
+║    - ONLY after sauron verification passes                       ║
+║                                                                   ║
 ║  Skipping ANY phase = Rule violation                             ║
 ╚══════════════════════════════════════════════════════════════════╝
 ```
@@ -71,6 +91,8 @@ Run manager agents 5 times to catch all issues:
 □ All agent counts match (CLAUDE.md, actual .md files in .claude/agents/)
 □ All agent frontmatter valid (required fields present)
 □ All skill references in agents exist in .claude/skills/
+□ All memory field values are valid (user | project | local)
+□ Memory distribution matches CLAUDE.md (24 project + 3 user + 7 none)
 □ All routing skill patterns updated
 ```
 
@@ -91,6 +113,7 @@ Run manager agents 5 times to catch all issues:
 □ No orphaned agents (not referenced by any routing skill)
 □ No circular references
 □ All skill references in agent frontmatter are valid
+□ All agent memory fields use valid scope values
 ```
 
 ### Deep Round 3: Philosophy Compliance
@@ -111,6 +134,7 @@ Phase 1: Manager Verification (5x)
   □ Round 3: mgr-supplier:audit (re-verify)
   □ Round 4: mgr-sync-checker:check + mgr-updater:docs
   □ Round 5: Final count verification
+  □ mgr-claude-code-bible:verify (official spec compliance)
 
 Phase 2: Deep Review (3x)
   □ Round 1: Workflow alignment
@@ -125,6 +149,11 @@ Phase 3: Fix Issues
 Phase 4: Commit
   □ Delegate to mgr-gitnerd
   □ Meaningful commit message
+
+Phase 5: Push
+  □ mgr-sauron:watch passed (MANDATORY)
+  □ Delegate push to mgr-gitnerd
+  □ Verify push succeeded
 ```
 
 ## When to Run Full Verification
@@ -173,10 +202,16 @@ Content Change Made
                 ▼
 ┌───────────────────────────────┐
 │  Phase 4: Commit via mgr-gitnerd │
+└───────────────┬───────────────┘
+                │
+                ▼
+┌───────────────────────────────┐
+│  Phase 5: Push via mgr-gitnerd  │
+│  (ONLY after sauron passes)    │
 └───────────────────────────────┘
 ```
 
-## Self-Check Before Commit
+## Self-Check Before Commit and Push
 
 ```
 ╔══════════════════════════════════════════════════════════════════╗
@@ -190,6 +225,18 @@ Content Change Made
 ║                                                                   ║
 ║  If NO to ANY → DO NOT COMMIT                                    ║
 ╚══════════════════════════════════════════════════════════════════╝
+
+╔══════════════════════════════════════════════════════════════════╗
+║  BEFORE PUSHING, ASK YOURSELF:                                   ║
+║                                                                   ║
+║  1. Did mgr-sauron:watch complete successfully?                  ║
+║  2. Were ALL issues from sauron verification fixed?              ║
+║  3. Am I delegating to mgr-gitnerd for the push?                 ║
+║                                                                   ║
+║  If NO to ANY → DO NOT PUSH                                      ║
+║                                                                   ║
+║  SAURON VERIFICATION IS MANDATORY FOR ALL PUSHES.                ║
+╚══════════════════════════════════════════════════════════════════╝
 ```
 
 ## Quick Verification Commands
@@ -215,6 +262,14 @@ done
 grep -c "agent_patterns:" .claude/skills/secretary-routing/SKILL.md
 grep -c "agent_patterns:" .claude/skills/dev-lead-routing/SKILL.md
 grep -c "agent_patterns:" .claude/skills/qa-lead-routing/SKILL.md
+
+# Memory field validation
+for f in .claude/agents/*.md; do
+  mem=$(grep "^memory:" "$f" | awk '{print $2}')
+  if [ -n "$mem" ] && [ "$mem" != "project" ] && [ "$mem" != "user" ] && [ "$mem" != "local" ]; then
+    echo "INVALID MEMORY SCOPE in $f: $mem"
+  fi
+done
 ```
 
 ## Integration with Other Rules

package/templates/.claude/rules/MUST-tool-identification.md

@@ -156,6 +156,29 @@ R008 extends R007 for tool operations:
 [Done] Agent created successfully
 ```
 
+## Task Tool Display Format
+
+When spawning subagents via the Task tool, the display MUST use the actual `subagent_type` parameter:
+
+```
+Task(subagent_type):model → description
+```
+
+### Examples
+
+```
+[secretary][opus] → Spawning parallel agents:
+  [1] Task(lang-golang-expert):sonnet → Go code review
+  [2] Task(lang-python-expert):sonnet → Python code review
+  [3] Task(general-purpose):haiku → File validation
+```
+
+### Rules
+
+- `subagent_type` MUST match the actual Task tool parameter value
+- Custom/invented names are NOT allowed (e.g., `code-review:sonnet` is WRONG)
+- Model specification shows cost/performance intent
+
 ## Benefits
 
 1. **Debugging**: Know which agent performed which operation

package/templates/.claude/rules/SHOULD-hud-statusline.md

@@ -46,10 +46,10 @@ Display real-time status information during agent operations for improved visibi
 ─── [Agent] secretary | [Progress] 0/4 | [Parallel] 4 ───
 
 Instances:
-  [1] readme-update:sonnet
-  [2] code-review:haiku
-  [3] architecture:opus
-  [4] validation:haiku
+  [1] Task(general-purpose):sonnet → README update
+  [2] Task(lang-golang-expert):haiku → Code review
+  [3] Task(Explore):opus → Architecture analysis
+  [4] Task(general-purpose):haiku → Validation
 ```
 
 ### Completion
@@ -85,12 +85,28 @@ echo "─── [Agent] $AGENT | [Progress] $PROGRESS ───" >&2
 
 ## Hook Usage
 
-The HUD statusline hook can be triggered via:
+The HUD statusline is implemented as an inline hook in `.claude/hooks/hooks.json` (PreToolUse → Task matcher).
 
-```bash
-~/.claude/hooks/hud/update-status.sh <agent> [progress] [parallel_count]
+The hook automatically displays subagent details when the Task tool is used:
+
+```
+─── [Spawn] {subagent_type}:{model} | {description} ───
+─── [Resume] {subagent_type}:{model} | {description} ───
+```
 
-# Examples
-~/.claude/hooks/hud/update-status.sh "mgr-creator" "1/3"
-~/.claude/hooks/hud/update-status.sh "secretary" "0/4" "4"
+### Examples
+
+```
+─── [Spawn] mgr-gitnerd:sonnet | Commit and push ───
+─── [Spawn] lang-golang-expert:sonnet | Review Go code ───
+─── [Spawn] mgr-creator:sonnet | Create new agent ───
+─── [Resume] mgr-gitnerd:sonnet | Continue push ───
 ```
+
+### Fields Displayed
+
+| Field | Source | Purpose |
+|-------|--------|---------|
+| `subagent_type` | Task tool parameter | Which agent is running |
+| `model` | Task tool parameter | Which model (opus/sonnet/haiku) |
+| `description` | Task tool parameter | What the agent is doing (max 40 chars) |