oh-my-customcodex 0.1.0

package/templates/.claude/skills/memory-recall/SKILL.md

@@ -0,0 +1,172 @@
+---
+name: memory-recall
+description: Search and recall memories from claude-mem
+scope: core
+argument-hint: "<query> [--recent] [--limit <n>]"
+user-invocable: true
+---
+
+# Memory Recall Skill
+
+Search and recall relevant memories from claude-mem using semantic search.
+
+## Parameters
+
+| Name | Type | Required | Description |
+|------|------|----------|-------------|
+| query | string | yes* | Search query (*not required with --recent) |
+
+## Options
+
+```
+--recent, -r     Recall most recent memories
+--limit, -l      Maximum results to return (default: 5)
+--verbose, -v    Show full memory content
+--date, -d       Filter by date (YYYY-MM-DD)
+```
+
+## Workflow
+
+```
+1. Build semantic query
+   ├── Prefix with "my-project"
+   ├── Add user query terms
+   └── Include date if specified
+
+2. Search claude-mem
+   └── chroma_query_documents
+
+3. Format results
+   ├── Sort by relevance score
+   ├── Show summary for each
+   └── Provide memory IDs for detailed access
+
+4. Present to user
+```
+
+## Query Tips
+
+```yaml
+effective_queries:
+  - "authentication flow" (topic-based)
+  - "2025-01-20 oauth" (temporal + topic)
+  - "decision jwt" (type + topic)
+  - "agent creation workflow" (semantic intent)
+
+avoid:
+  - Single generic words like "code" or "task"
+  - Very long queries (keep it focused)
+  - Special characters or complex syntax
+```
+
+## Retrieval Strategy
+
+### Recall-Precision Tradeoff
+
+Default bias: **recall > precision** — it is easier to filter out irrelevant results (false positives) than to recover missed information (false negatives).
+
+| Task Type | Bias | Recommended --limit | Rationale |
+|-----------|------|--------------------:|-----------|
+| Debugging / Investigation | High recall (16:1) | 10-15 | Cast wide net, prune later |
+| Decision reference | Balanced | 5 (default) | Moderate breadth with manageable noise |
+| Specific fact lookup | High precision | 3 | Narrow, targeted retrieval |
+
+### Guidelines
+
+- **Over-retrieve, then filter**: When uncertain, request more results and discard irrelevant ones in post-processing
+- **Narrow progressively**: Start broad, refine query only if results are noisy — avoid starting too narrow
+- **Combine temporal + semantic**: Add date filters (`--date`) to semantic queries for better precision without sacrificing recall
+
+## Output Format
+
+### Basic Search
+```
+[sys-memory-keeper:recall authentication]
+
+Searching memories for: "my-project authentication"
+
+Found 3 relevant memories:
+
+[1] mem_abc123 (Score: 0.95)
+    Date: 2025-01-20
+    Summary: Implemented OAuth flow with Google provider
+    Tags: [authentication, oauth, google]
+
+[2] mem_def456 (Score: 0.87)
+    Date: 2025-01-18
+    Summary: JWT token decision - RS256 algorithm
+    Tags: [authentication, decision, jwt]
+
+[3] mem_ghi789 (Score: 0.82)
+    Date: 2025-01-15
+    Summary: Authentication architecture discussion
+    Tags: [authentication, architecture, planning]
+
+Use "sys-memory-keeper:recall --verbose" for full content.
+```
+
+### Verbose Output
+```
+[sys-memory-keeper:recall authentication --verbose]
+
+Searching memories for: "my-project authentication"
+
+[1] mem_abc123 (Score: 0.95)
+    Date: 2025-01-20
+    Tags: [authentication, oauth, google]
+
+    Content:
+    ## Session Summary
+    Date: 2025-01-20
+
+    ### Tasks Completed
+    - Implemented OAuth flow with Google provider
+    - Added callback handler for OAuth response
+    - Created user session management
+
+    ### Decisions Made
+    - Use passport.js for OAuth handling
+      Rationale: Well-maintained, good documentation
+
+    ---
+
+[2] mem_def456 (Score: 0.87)
+    ...
+```
+
+### Recent Memories
+```
+[sys-memory-keeper:recall --recent]
+
+Fetching recent memories for: my-project
+
+[1] mem_xyz999 (Score: 1.00)
+    Date: 2025-01-24
+    Summary: Memory system Phase 1 implementation
+    Tags: [session, memory, phase1]
+
+[2] mem_xyz888 (Score: 0.98)
+    Date: 2025-01-23
+    Summary: Agent identification rules update
+    Tags: [session, rules, enforcement]
+
+...
+```
+
+### No Results
+```
+[sys-memory-keeper:recall "nonexistent topic"]
+
+Searching memories for: "my-project nonexistent topic"
+
+No memories found matching your query.
+
+Suggestions:
+- Try different keywords
+- Use broader search terms
+- Check available memories with "sys-memory-keeper:recall --recent"
+```
+
+## Related
+
+- memory-save - Save current context

package/templates/.claude/skills/memory-save/SKILL.md

@@ -0,0 +1,128 @@
+---
+name: memory-save
+description: Save current session context to claude-mem
+scope: core
+argument-hint: "[--tags <tags>] [--include-code]"
+disable-model-invocation: true
+user-invocable: true
+---
+
+# Memory Save Skill
+
+Save current session context to claude-mem for persistence across context compaction.
+
+## Options
+
+```
+--tags, -t       Additional tags for the memory
+--include-code   Include code changes in the save
+--summary, -s    Custom summary (otherwise auto-generated)
+--verbose, -v    Show detailed save information
+```
+
+## Workflow
+
+```
+1. Collect session context
+   ├── Tasks completed
+   ├── Decisions made
+   ├── Open items
+   └── Code changes (if --include-code)
+
+2. Format with metadata
+   ├── project: my-project
+   ├── session: {date}-{uuid}
+   ├── tags: [session, ...user_tags]
+   └── created_at: {timestamp}
+
+3. Store in claude-mem
+   └── chroma_add_documents
+
+4. Report result
+```
+
+## Storage Format
+
+```yaml
+project: my-project
+session: {date}-{uuid}
+tags: [session, task, decision]
+content:
+  summary: Brief description of session context
+  tasks_completed: List of completed tasks
+  decisions: Key decisions made
+  open_items: Unfinished work
+```
+
+## Output Format
+
+### Success
+```
+[sys-memory-keeper:save]
+
+Saving session context...
+
+Context collected:
+  Tasks: 3 completed
+  Decisions: 2 recorded
+  Open items: 1 pending
+
+Metadata:
+  Project: my-project
+  Session: 2025-01-24-a1b2c3d4
+  Tags: [session, task, decision]
+
+[Done] Session context saved successfully.
+Memory ID: mem_abc123
+```
+
+### With Tags
+```
+[sys-memory-keeper:save --tags "authentication,oauth"]
+
+Saving session context...
+
+Metadata:
+  Project: my-project
+  Session: 2025-01-24-a1b2c3d4
+  Tags: [session, task, decision, authentication, oauth]
+
+[Done] Session context saved successfully.
+Memory ID: mem_abc123
+```
+
+### Verbose
+```
+[sys-memory-keeper:save --verbose]
+
+Collecting session context...
+
+Tasks Completed:
+  1. Implemented OAuth flow
+  2. Added JWT token validation
+  3. Created authentication middleware
+
+Decisions Made:
+  1. Use RS256 for JWT signing
+     Rationale: Better security for distributed systems
+  2. Token expiry: 1 hour
+     Rationale: Balance security and user experience
+
+Open Items:
+  1. Refresh token implementation
+     Status: In progress
+
+Saving to claude-mem...
+
+Document content:
+  ## Session Summary
+  Date: 2025-01-24
+  ...
+
+[Done] Session context saved.
+Memory ID: mem_abc123
+```
+
+## Related
+
+- memory-recall - Search and recall memories

package/templates/.claude/skills/model-escalation/SKILL.md

@@ -0,0 +1,62 @@
+---
+name: model-escalation
+description: Advisory model escalation based on task outcome tracking
+scope: core
+user-invocable: false
+---
+
+# Model Escalation Skill
+
+Tracks task outcomes and advises model upgrades when failures are detected. **Advisory-only** — the orchestrator makes the final decision (R010).
+
+## Escalation Path
+
+```
+haiku → sonnet → opus
+```
+
+## Trigger Conditions
+
+| Condition | Action |
+|-----------|--------|
+| 2+ failures with same model for same agent type | Advise escalation |
+| 3+ consecutive failures across any agent type | Advise global escalation |
+| Sustained success after escalation | Advise de-escalation |
+
+## Thresholds
+
+| Parameter | Default | Description |
+|-----------|---------|-------------|
+| `failure_threshold` | 2 | Failures before escalation advisory |
+| `consecutive_threshold` | 3 | Consecutive failures for global advisory |
+| `cooldown_tasks` | 5 | Successes before de-escalation advisory |
+
+## Cost Guard
+
+- Advisory includes estimated cost multiplier
+- De-escalation suggested after sustained success at higher tier
+- Cost tracked per session via PPID-scoped temp file
+
+## Architecture
+
+```
+PostToolUse (Task) → task-outcome-recorder.sh
+  Records: agent_type, model, success/failure, error_summary
+  Storage: /tmp/.codex-task-outcomes-$PPID (JSON lines, max 50)
+
+PreToolUse (Task) → model-escalation-advisor.sh
+  Reads outcomes → counts failures → advises escalation via stderr
+  Advisory only — never blocks, never modifies tool input
+```
+
+## Advisory Format
+
+```
+--- [Model Escalation Advisory] ---
+  Agent type: {agent_type}
+  Current model: {current_model}
+  Recent failures: {count}/{threshold}
+  ⚡ Recommended: Escalate to {next_model}
+  Cost impact: {multiplier} per task
+---
+```

package/templates/.claude/skills/monitoring-setup/SKILL.md

@@ -0,0 +1,215 @@
+---
+name: omcodex:monitoring-setup
+description: Enable/disable OpenTelemetry console monitoring for GPT Codex + OMX usage tracking
+scope: package
+argument-hint: "[enable|disable|status]"
+user-invocable: true
+---
+
+# Monitoring Setup Skill
+
+Enable or disable OpenTelemetry console monitoring. When enabled, the harness outputs usage metrics (cost, tokens, sessions, LOC, commits, PRs, active time) and events (tool results, API requests) to the terminal.
+
+## Natural Language Triggers
+
+This skill activates when the user mentions any of:
+- Korean: "모니터링", "텔레메트리", "사용량 추적", "메트릭", "모니터링 켜줘", "텔레메트리 활성화"
+- English: "monitoring", "telemetry", "usage tracking", "metrics", "enable monitoring"
+- Combined with actions: "켜", "끄", "활성화", "비활성화", "설정", "enable", "disable", "setup"
+
+## Commands
+
+### enable (default)
+
+1. Read `.codex/settings.local.json` (create if not exists)
+2. Add or update `env` field with:
+   ```json
+   {
+     "env": {
+       "CLAUDE_CODE_ENABLE_TELEMETRY": "1",
+       "OTEL_METRICS_EXPORTER": "console",
+       "OTEL_LOGS_EXPORTER": "console"
+     }
+   }
+   ```
+3. Preserve all existing settings
+4. Report to user:
+   ```
+   [Done] OpenTelemetry Console Monitoring enabled
+
+   Configured in: .codex/settings.local.json
+   Metrics: sessions, cost, tokens, LOC, commits, PRs, active time
+   Events: tool results, API requests, tool decisions
+
+   Note: Takes effect on next session restart.
+   To disable: /monitoring-setup disable
+   ```
+
+### disable
+
+1. Read `.codex/settings.local.json`
+2. Remove OTel-related keys from `env`:
+   - `CLAUDE_CODE_ENABLE_TELEMETRY`
+   - `OTEL_METRICS_EXPORTER`
+   - `OTEL_LOGS_EXPORTER`
+3. If `env` object becomes empty, remove `env` field entirely
+4. Report:
+   ```
+   [Done] OpenTelemetry Monitoring disabled
+
+   Removed from: .codex/settings.local.json
+   Takes effect on next session restart.
+   ```
+
+### status
+
+1. Read `.codex/settings.local.json`
+2. Check for OTel env vars
+3. Report current state:
+   ```
+   [Monitoring Status]
+   ├── Enabled: Yes/No
+   ├── Metrics exporter: console / otlp / none
+   ├── Logs exporter: console / otlp / none
+   └── Config: .codex/settings.local.json
+   ```
+
+## Implementation Notes
+
+- `settings.local.json` is NOT git-tracked (local to user)
+- Each user enables monitoring independently
+- No infrastructure required for console mode
+- Metrics appear in stderr during GPT Codex + OMX session execution
+- Default export interval: 60s for metrics, 5s for events
+
+## Available Metrics
+
+| Metric | Description | Unit |
+|--------|-------------|------|
+| `claude_code.session.count` | CLI sessions started | count |
+| `claude_code.cost.usage` | Session cost | USD |
+| `claude_code.token.usage` | Tokens used (input/output/cache) | tokens |
+| `claude_code.lines_of_code.count` | Code lines modified (added/removed) | count |
+| `claude_code.commit.count` | Git commits created | count |
+| `claude_code.pull_request.count` | Pull requests created | count |
+| `claude_code.active_time.total` | Active usage time | seconds |
+
+## Available Events
+
+| Event | Description |
+|-------|-------------|
+| `claude_code.tool_result` | Tool execution results with duration |
+| `claude_code.api_request` | API request details with cost/tokens |
+| `claude_code.api_error` | API error details |
+| `claude_code.tool_decision` | Tool accept/reject decisions |
+| `claude_code.user_prompt` | User prompt metadata (content redacted by default) |
+
+## Upgrade Path
+
+For production monitoring, upgrade from console to OTLP:
+
+```bash
+# In settings.local.json env:
+OTEL_METRICS_EXPORTER=otlp
+OTEL_LOGS_EXPORTER=otlp
+OTEL_EXPORTER_OTLP_PROTOCOL=grpc
+OTEL_EXPORTER_OTLP_ENDPOINT=http://localhost:4317
+```
+
+## Advanced OTel Configuration
+
+### Additional Metrics
+
+| Metric | Description | Unit |
+|--------|-------------|------|
+| `code_edit_tool.decision` | Edit tool accept/reject decisions | count |
+
+### Exporter Configuration
+
+```json
+{
+  "env": {
+    "OTEL_METRICS_EXPORTER": "otlp",
+    "OTEL_LOGS_EXPORTER": "otlp",
+    "OTEL_EXPORTER_OTLP_PROTOCOL": "grpc",
+    "OTEL_EXPORTER_OTLP_ENDPOINT": "http://localhost:4317",
+    "OTEL_RESOURCE_ATTRIBUTES": "service.name=claude-code,service.version=2.1.85"
+  }
+}
+```
+
+### Cardinality Controls
+
+| Variable | Description | Default |
+|----------|-------------|---------|
+| `OTEL_LOG_TOOL_DETAILS` | Include tool input/output in logs | `false` |
+| `OTEL_METRICS_INCLUDE_TOOL_NAME` | Include tool name dimension | `true` |
+| `OTEL_METRICS_INCLUDE_MODEL` | Include model dimension | `true` |
+
+### Multi-Exporter Syntax
+
+```bash
+# Send metrics to both console and OTLP
+OTEL_METRICS_EXPORTER=console,otlp
+OTEL_LOGS_EXPORTER=console,otlp
+```
+
+### Prometheus Exporter
+
+```bash
+OTEL_METRICS_EXPORTER=prometheus
+OTEL_EXPORTER_PROMETHEUS_PORT=9464
+```
+
+## Data Privacy
+
+Environment variables to control data collection and telemetry:
+
+| Variable | Description |
+|----------|-------------|
+| `DISABLE_TELEMETRY` | Disable all telemetry collection |
+| `DISABLE_ERROR_REPORTING` | Disable error reporting to Anthropic |
+| `CLAUDE_CODE_DISABLE_NONESSENTIAL_TRAFFIC` | Disable non-essential network traffic |
+| `CLAUDE_CODE_DISABLE_FEEDBACK_SURVEY` | Disable feedback survey prompts |
+| `DO_NOT_TRACK` | Standard DNT signal |
+
+### Enterprise Configuration
+
+```json
+{
+  "env": {
+    "DISABLE_TELEMETRY": "1",
+    "DISABLE_ERROR_REPORTING": "1",
+    "CLAUDE_CODE_DISABLE_NONESSENTIAL_TRAFFIC": "1"
+  }
+}
+```
+
+## HTTP-Level Inspection (Optional)
+
+For deeper payload-level debugging beyond aggregated metrics, [Claude Inspector](https://github.com/kangraemin/claude-inspector) provides MITM proxy inspection of Claude Code HTTP traffic.
+
+| Aspect | OTel Monitoring (this skill) | Claude Inspector |
+|--------|------------------------------|-----------------|
+| Layer | Application (hooks, stdout) | HTTP (MITM proxy) |
+| Metrics | Aggregated (cost, tokens, duration) | Per-request payload breakdown |
+| Cache visibility | Not available | Prompt Cache hit/miss rates |
+| Sub-agent view | Summary via hooks | Full parent vs sub-agent context comparison |
+| Setup | Built-in (hooks + statusline) | External tool (Homebrew on macOS) |
+
+### When to Use
+
+- **OTel monitoring**: Daily operations, cost tracking, performance trends
+- **Claude Inspector**: Debugging specific payload issues, measuring AGENTS.md token impact, verifying ecomode (R013) effectiveness, profiling sub-agent context inheritance
+
+### Setup
+
+```bash
+# macOS
+brew install kangraemin/tap/claude-inspector
+
+# Run proxy
+claude-inspector
+```
+
+Claude Inspector is external to oh-my-customcodex and does not require any project configuration changes.

package/templates/.claude/skills/multi-model-verification/SKILL.md

@@ -0,0 +1,130 @@
+---
+name: multi-model-verification
+description: Parallel code verification using multiple models with severity classification
+scope: core
+version: 1.0.0
+user-invocable: false
+---
+
+# Multi-Model Verification
+
+Parallel code verification using multiple AI models for comprehensive quality assurance. Each model focuses on a different aspect, and results are aggregated with severity classification.
+
+## Background
+
+Inspired by Pi Coding Agent Workflow Extension's multi-model verification pattern. Instead of a single reviewer, multiple models verify code simultaneously, each with specialized focus areas.
+
+## Verification Roles
+
+| Model | Role | Focus Areas |
+|-------|------|-------------|
+| `opus` | Architecture Reviewer | Design patterns, separation of concerns, extensibility, security architecture |
+| `sonnet` | Code Quality Reviewer | Logic correctness, error handling, edge cases, performance patterns |
+| `haiku` | Style & Convention Reviewer | Naming conventions, formatting, documentation, code organization |
+
+## Severity Classification
+
+| Severity | Description | Action Required |
+|----------|-------------|-----------------|
+| CRITICAL | Bugs, security vulnerabilities, data loss risks | Must fix before merge |
+| WARNING | Code smells, suboptimal patterns, missing error handling | Should fix, justify if skipped |
+| INFO | Style suggestions, minor improvements, alternative approaches | Optional improvement |
+
+## Workflow
+
+### Prerequisites
+- Agent Teams enabled (`OMCODEX_AGENT_TEAMS=1`) for full parallel execution
+- Falls back to sequential Agent tool execution if Agent Teams unavailable
+
+### Execution Flow
+
+1. **Input**: File path(s) or diff to verify
+2. **Spawn Parallel Reviewers**:
+   - `Agent(opus)` → Architecture & design review
+   - `Agent(sonnet)` → Code quality & correctness review
+   - `Agent(haiku)` → Style & convention review
+3. **Collect Results**: Each reviewer returns findings with severity
+4. **Aggregate**: Merge and deduplicate findings
+5. **Report**: Unified report sorted by severity
+
+### Agent Teams Mode
+
+When Agent Teams is available, create a verification team:
+
+```
+TeamCreate("verification-team")
+├── architect-reviewer (opus) → Architecture review
+├── quality-reviewer (sonnet) → Code quality review
+└── style-reviewer (haiku) → Style review
+```
+
+Members communicate findings via SendMessage for cross-cutting concerns.
+
+### Agent Tool Fallback
+
+When Agent Teams is unavailable, spawn parallel agents:
+
+```
+[1] Agent(general-purpose):opus → Architecture review
+[2] Agent(general-purpose):sonnet → Code quality review
+[3] Agent(general-purpose):haiku → Style & convention review
+```
+
+## Output Format
+
+```
+## Multi-Model Verification Report
+
+### Summary
+- Files reviewed: {count}
+- Findings: {critical} CRITICAL, {warning} WARNING, {info} INFO
+- Reviewers: opus (architecture), sonnet (quality), haiku (style)
+
+### CRITICAL
+[opus] {file}:{line} — {description}
+[sonnet] {file}:{line} — {description}
+
+### WARNING
+[sonnet] {file}:{line} — {description}
+[haiku] {file}:{line} — {description}
+
+### INFO
+[haiku] {file}:{line} — {description}
+
+### Consensus
+Issues flagged by 2+ reviewers:
+- {file}:{line} — {description} (flagged by: opus, sonnet)
+```
+
+## Integration
+
+- Works with `dev-review` skill as an enhanced verification mode
+- Integrates with `structured-dev-cycle` skill at the "Verify Implementation" stage
+- Compatible with R009 (parallel execution) and R018 (Agent Teams) rules
+
+## Configuration
+
+| Parameter | Default | Description |
+|-----------|---------|-------------|
+| `models` | `[opus, sonnet, haiku]` | Models to use for verification |
+| `severity_filter` | `all` | Minimum severity to report (`critical`, `warning`, `all`) |
+| `consensus_threshold` | `2` | Number of reviewers needed for consensus flag |
+| `include_suggestions` | `true` | Include INFO-level suggestions |
+
+## When to Use
+
+| Scenario | Recommended |
+|----------|-------------|
+| Pre-merge review of critical code | Yes |
+| Architecture changes | Yes |
+| Security-sensitive code | Yes |
+| Simple formatting changes | No (use single reviewer) |
+| Rapid prototyping | No (overhead too high) |
+
+## Cost Awareness
+
+Multi-model verification uses 3x the tokens of a single review. Reserve for:
+- Critical path code changes
+- Security-sensitive modifications
+- Architecture decisions
+- Pre-release verification