oh-my-customcodex 0.1.0

package/templates/.claude/skills/npm-audit/SKILL.md

@@ -0,0 +1,74 @@
+---
+name: omcodex:npm-audit
+description: Audit npm dependencies for security and updates
+scope: package
+argument-hint: "[--fix] [--production]"
+user-invocable: true
+---
+
+# NPM Audit Skill
+
+Audit npm dependencies for security vulnerabilities and outdated packages.
+
+## Options
+
+```
+--fix            Automatically fix vulnerabilities where possible
+--production     Only audit production dependencies
+--json           Output in JSON format
+```
+
+## Workflow
+
+```
+1. Run npm audit for security vulnerabilities
+2. Analyze vulnerability severity
+3. Check for outdated dependencies
+4. Generate health report
+5. Suggest remediation steps
+```
+
+## Output Format
+
+### Success
+```
+[NPM Audit] package-name
+
+Security:
+  Critical: 0
+  High: 0
+  Moderate: 2
+  Low: 1
+
+Outdated:
+  Major updates: 3
+  Minor updates: 5
+  Patch updates: 12
+
+Status: Needs attention (2 moderate vulnerabilities)
+
+Recommendations:
+  1. npm update lodash
+  2. npm update axios
+```
+
+### Failure
+```
+[NPM Audit] Failed
+
+Error: {error_message}
+Hint: Ensure package-lock.json exists
+```
+
+## Examples
+
+```bash
+# Full dependency audit with report
+npm-audit
+
+# Audit and fix vulnerabilities
+npm-audit --fix
+
+# Audit only production dependencies
+npm-audit --production
+```

package/templates/.claude/skills/npm-publish/SKILL.md

@@ -0,0 +1,65 @@
+---
+name: omcodex:npm-publish
+description: Publish package to npm registry with pre-checks
+scope: package
+argument-hint: "[--tag <tag>] [--dry-run]"
+disable-model-invocation: true
+user-invocable: true
+---
+
+# NPM Publish Skill
+
+Publish package to npm registry with comprehensive pre-publish checks and validation.
+
+## Options
+
+```
+--tag            npm dist-tag (default: latest)
+--dry-run        Run all checks without publishing
+--skip-tests     Skip test execution
+```
+
+## Workflow
+
+```
+1. Validate package.json configuration
+2. Check version against registry
+3. Run pre-publish checks (tests, lint, build)
+4. Execute npm pack (dry-run)
+5. Publish with appropriate tag
+6. Verify publication success
+```
+
+## Output Format
+
+### Success
+```
+[NPM Publish] package-name@1.2.3
+
+Pre-checks: All passed
+Registry: https://registry.npmjs.org
+Tag: latest
+
+Package published successfully.
+```
+
+### Failure
+```
+[NPM Publish] Failed
+
+Error: {error_message}
+Suggested fix: {fix_suggestion}
+```
+
+## Examples
+
+```bash
+# Publish to npm with all checks
+npm-publish
+
+# Publish to beta tag
+npm-publish --tag beta
+
+# Validate without publishing
+npm-publish --dry-run
+```

package/templates/.claude/skills/npm-version/SKILL.md

@@ -0,0 +1,104 @@
+---
+name: omcodex:npm-version
+description: Manage semantic versions for npm packages
+scope: package
+argument-hint: "<major|minor|patch> [--no-tag] [--no-commit]"
+disable-model-invocation: true
+user-invocable: true
+---
+
+# NPM Version Management Skill
+
+Manage semantic versions for npm packages with automatic changelog and git integration.
+
+## Arguments
+
+```
+major            Bump major version (x.0.0)
+minor            Bump minor version (0.x.0)
+patch            Bump patch version (0.0.x)
+```
+
+## Options
+
+```
+--no-tag         Skip git tag creation
+--no-commit      Skip commit creation (only update files)
+```
+
+## Workflow
+
+```
+1. Analyze current version from package.json
+2. Determine version bump type
+3. Update package.json version field
+4. Update CHANGELOG.md if exists
+5. Create version commit
+6. Create git tag (optional)
+```
+
+## Output Format
+
+### Success
+```
+[NPM Version] package-name
+
+Previous: 1.2.3
+Current:  1.2.4
+
+Changes:
+  - package.json updated
+  - Commit: "chore: bump version to 1.2.4"
+  - Tag: v1.2.4
+```
+
+### Failure
+```
+[NPM Version] Failed
+
+Error: {error_message}
+Hint: Ensure clean git working directory
+```
+
+## Examples
+
+```bash
+# Bump patch version (1.2.3 -> 1.2.4)
+npm-version patch
+
+# Bump minor version (1.2.3 -> 1.3.0)
+npm-version minor
+
+# Bump major version (1.2.3 -> 2.0.0)
+npm-version major
+
+# Update version without creating git tag
+npm-version patch --no-tag
+```
+
+## Release Branch Integration
+
+When working with `auto-tag.yml` (automatic tag creation on release PR merge):
+
+1. `.npmrc` has `git-tag-version=false` — prevents local tag creation during `npm version`
+2. `auto-tag.yml` creates the tag on the **merge commit** when a `release/*` PR is merged to `develop`
+3. Do NOT manually push tags — let the CI workflow handle tag creation
+
+### Release Workflow
+
+```
+1. Create release branch: release/vX.Y.Z
+2. Run version bump (npm version / manual edit)  ← no local tag created
+3. Build dist, commit, push
+4. Create PR → merge to develop
+5. auto-tag.yml creates vX.Y.Z tag on merge commit  ← correct tag target
+6. release.yml triggers on tag → GitHub Release + npm publish
+```
+
+### Troubleshooting
+
+If a tag already exists on remote (from a previous failed attempt):
+```bash
+git push origin :refs/tags/vX.Y.Z   # delete remote tag
+# Then re-merge or let auto-tag.yml handle it
+```

package/templates/.claude/skills/omcodex-auto-improve/SKILL.md

@@ -0,0 +1,136 @@
+---
+name: omcodex:auto-improve
+description: Apply verified improvement suggestions from eval-core analysis to omcodex configuration
+scope: harness
+user-invocable: true
+effort: high
+---
+
+# /omcodex:auto-improve — Automated Improvement Workflow
+
+## Purpose
+
+Reads improvement suggestions from eval-core analysis, lets the user select which to apply, applies changes in an isolated worktree with sauron verification, and creates a PR for review.
+
+## Usage
+
+```
+/omcodex:auto-improve              # Interactive selection from pending suggestions
+```
+
+## Prerequisites
+
+- eval-core analysis data exists (run `/omcodex:improve-report` first if empty)
+- Pending improvement suggestions in `proposed` status
+
+## Workflow
+
+### Step 1: Read Suggestions
+
+1. Run `bun run packages/eval-core/src/cli/index.ts analyze --format json --save` via Bash
+2. Parse JSON output for improvement suggestions
+3. If no suggestions: display "No improvement suggestions available" and exit
+
+### Step 2: Display & Select
+
+Display numbered list:
+```
+[Auto-Improve] Available suggestions:
+  1. [HIGH] agent:lang-golang-expert — Escalate model sonnet→opus (3 failures in 5 uses)
+  2. [MED]  routing:dev-lead-routing — Add Flutter keyword mapping (2 routing misses)
+  3. [LOW]  skill:systematic-debugging — Add timeout guard (1 timeout in 10 uses)
+
+Select items: [1,2,3] / "all" / "cancel"
+```
+
+**Self-reference filter**: Exclude items where targetName matches:
+- `omcodex:auto-improve`, `auto-improve`
+- `pipeline-guards`, `evaluator-optimizer`
+- Any item targeting this skill itself
+
+### Step 3: Approve (State Transition)
+
+For each selected item:
+1. Call eval-core API: transition `proposed` → `approved`
+2. Display: `[Approved] {N} items selected for application`
+
+### Step 4: Worktree Isolation
+
+- Use `EnterWorktree` tool with name `auto-improve-{YYYYMMDD}`
+- Creates isolated branch from HEAD
+
+### Step 5: Apply Changes
+
+Map each approved item to the appropriate subagent by `targetType`:
+
+| targetType | Agent | Action |
+|------------|-------|--------|
+| agent | mgr-creator | Modify agent frontmatter/body |
+| skill | Matching domain expert | Revise skill SKILL.md |
+| routing | general-purpose | Update routing patterns |
+| model-escalation | general-purpose | Update model field in agent frontmatter |
+
+Spawn agents in parallel (max 4 per R009). Each agent receives:
+- Action description and evidence data
+- Target file path
+- Specific modification instructions
+
+### Step 6: Verification
+
+1. Delegate to mgr-sauron: full R017 verification
+2. If **PASS**: proceed to Step 7
+3. If **FAIL**: display failures, offer options:
+   - `fix` → re-apply with sauron feedback (max 2 cycles)
+   - `reject` → transition all to `rejected`, ExitWorktree(remove)
+   - `manual` → keep worktree for user inspection
+
+### Step 7: PR & Finalize
+
+1. Delegate to mgr-gitnerd: commit + create PR
+   - Title: `chore(auto-improve): apply {N} improvement suggestions`
+   - Body: table of applied items with evidence
+2. Transition all items to `applied` with `appliedAt` timestamp and PR URL
+3. `ExitWorktree(action: "keep")` — keep branch for PR
+4. Display PR URL to user
+
+## Safety Guards
+
+| Guard | Implementation |
+|-------|---------------|
+| Self-reference prevention | Blocklist filter in Step 2 |
+| User approval gate | Step 2 interactive selection |
+| Worktree isolation | Step 4 EnterWorktree |
+| Sauron verification | Step 6 mandatory pass |
+| PR-based merge | Step 7 — no direct push to develop |
+| Max items per run | 20 default, 50 hard cap |
+| Max fix cycles | 2 retries before rejection |
+| Rollback | `git revert` via mgr-gitnerd post-merge |
+
+## Error Handling
+
+| Scenario | Action |
+|----------|--------|
+| No suggestions available | Display message, exit |
+| User cancels selection | Exit, no state changes |
+| Sauron verification fails 2x | Reject all, cleanup worktree |
+| Agent application error | Mark individual item as rejected, continue others |
+| EnterWorktree fails | Report error, exit |
+
+## Display Format
+
+```
+[Auto-Improve] Starting improvement workflow
+├── Suggestions: {N} available ({high}H/{medium}M/{low}L confidence)
+├── Self-reference filtered: {count} items excluded
+└── Select items to apply: [1,2,3] or "all" or "cancel"
+
+[Auto-Improve] Applying {N} improvements in worktree
+├── Worktree: auto-improve-{date}
+├── Agents: {count} parallel
+└── Pipeline guards: max 20 items, 2 retry cycles
+
+[Auto-Improve] Verification
+├── Sauron: {PASS|FAIL}
+├── PR: #{number} created
+└── Status: {N} items → applied
+```

package/templates/.claude/skills/omcodex-feedback/SKILL.md

@@ -0,0 +1,205 @@
+---
+name: omcodex:feedback
+description: Submit feedback about oh-my-customcodex (supports anonymous submission)
+scope: harness
+user-invocable: true
+disable-model-invocation: true
+argument-hint: "[description or leave empty for interactive] [--anonymous]"
+---
+
+# Feedback Submitter
+
+Submit feedback about oh-my-customcodex (bugs, features, improvements, questions) directly from the CLI session. Supports anonymous submission with `[Anonymous Feedback]` title prefix when `--anonymous` flag is used.
+
+## Purpose
+
+Lowers the barrier for submitting feedback by allowing users to create GitHub issues — without leaving their terminal session. All feedback is filed to the `baekenough/oh-my-customcodex` repository.
+
+## Usage
+
+```
+# Inline feedback
+/omcodex:feedback HUD display is missing during parallel agent spawn
+
+# Anonymous submission
+/omcodex:feedback --anonymous Something feels off with the routing
+
+# Interactive (no arguments)
+/omcodex:feedback
+```
+
+## Workflow
+
+### Phase 1: Input Parsing
+
+Check for `--anonymous` flag in the arguments:
+- If `--anonymous` is present, set `ANONYMOUS=true` and strip the flag from the content
+- Otherwise, set `ANONYMOUS=false`
+
+If remaining arguments are provided:
+1. Analyze the content to auto-detect category (`bug`, `feature`, `improvement`, `question`)
+2. Use the content as the issue title (truncate to 80 chars if needed)
+3. Use the full content as the description body
+
+If no arguments (or only `--anonymous`):
+1. Ask the user for category using AskUserQuestion: `[bug / feature / improvement / question]`
+2. Ask for title and optional detailed description (combine into a single prompt when possible)
+
+### Phase 2: Route Decision
+
+Check environment and user intent:
+
+```bash
+# Check gh CLI availability
+command -v gh >/dev/null 2>&1 && GH_AVAILABLE=true || GH_AVAILABLE=false
+
+# Check gh authentication (only if gh is available)
+if [ "$GH_AVAILABLE" = "true" ]; then
+  gh auth status >/dev/null 2>&1 && GH_AUTHED=true || GH_AUTHED=false
+else
+  GH_AUTHED=false
+fi
+```
+
+**Route A**: `gh` available + authenticated
+- Use GitHub Issue creation (see Phase 4A)
+- If `--anonymous`: adds `[Anonymous Feedback]` prefix and `anonymous` label
+
+**Fallback**: `gh` NOT available or not authenticated
+- Save feedback locally and inform the user (see Phase 4D)
+
+### Phase 3: Environment Collection
+
+Collect environment info via Bash:
+
+```bash
+# omcodex version
+OMCODEX_VERSION=$(node -e "console.log(require('./package.json').version)" 2>/dev/null || echo "unknown")
+
+# Codex version
+CODEX_VERSION=$(codex --version 2>/dev/null || echo "unknown")
+
+# OS
+OS_INFO=$(uname -s 2>/dev/null || echo "unknown")
+
+# Project name
+PROJECT_NAME=$(basename "$(pwd)")
+
+# Build project context string
+PROJECT_CONTEXT="omcodex v${OMCODEX_VERSION}, Codex ${CODEX_VERSION}, ${OS_INFO}"
+```
+
+For anonymous submissions, do NOT include the project name. Offer to include project context as opt-in:
+- Ask: "Include environment info (version, OS) in the anonymous report? [Y/n]"
+- If declined, set `PROJECT_CONTEXT=""`
+
+### Phase 4A: GitHub Issue Creation (Route A — gh + authenticated)
+
+1. If `ANONYMOUS=true`, prepend `[Anonymous Feedback] ` to the title and add `anonymous` to the label list.
+
+2. Show the user a preview of the issue to be created:
+   ```
+   [Preview]
+   ├── Title: {title}
+   ├── Category: {category}
+   ├── Labels: feedback, {category-label}[, anonymous]
+   └── Repo: baekenough/oh-my-customcodex
+   ```
+3. Ask for confirmation before creating
+
+4. Ensure labels exist (defensive):
+   ```bash
+   gh label create feedback --description "User feedback via /omcodex:feedback" --color 0E8A16 --repo baekenough/oh-my-customcodex 2>/dev/null || true
+   # If anonymous, ensure the anonymous label exists
+   if [ "$ANONYMOUS" = "true" ]; then
+     gh label create anonymous --description "Anonymous feedback submission" --color C5DEF5 --repo baekenough/oh-my-customcodex 2>/dev/null || true
+   fi
+   ```
+
+5. Create the issue using `--body-file` for safe markdown handling:
+   ```bash
+   # Write body to temp file to avoid shell escaping issues
+   cat > /tmp/omcodex-feedback-body.md << 'FEEDBACK_EOF'
+   ## Feedback
+
+   **Category**: {category}
+   **Source**: omcodex CLI v{version}
+
+   ### Description
+   {user description}
+
+   ### Environment
+   - omcodex version: {omcodex_version}
+   - Codex version: {codex_version}
+   - OS: {os_info}
+   - Project: {project_name}
+
+   ---
+   *Submitted via `/omcodex:feedback`*
+   FEEDBACK_EOF
+
+   # Build label string
+   LABELS="feedback,${CATEGORY_LABEL}"
+   if [ "$ANONYMOUS" = "true" ]; then
+     LABELS="${LABELS},anonymous"
+   fi
+
+   # Create issue
+   gh issue create \
+     --repo baekenough/oh-my-customcodex \
+     --title "{title}" \
+     --label "$LABELS" \
+     --body-file /tmp/omcodex-feedback-body.md
+
+   # Clean up
+   rm -f /tmp/omcodex-feedback-body.md
+   ```
+
+6. If label creation fails AND issue creation fails due to labels, retry without labels as fallback
+
+7. Return the issue URL to the user
+
+### Phase 4D: Local Fallback (gh not available, not authenticated, or issue creation failed)
+
+```bash
+mkdir -p ~/.omcodex/feedback
+TIMESTAMP=$(date +%Y%m%dT%H%M%S)
+FEEDBACK_FILE=~/.omcodex/feedback/${TIMESTAMP}.json
+
+cat > "$FEEDBACK_FILE" << EOF
+{
+  "title": "$TITLE",
+  "body": "$BODY",
+  "feedback_type": "$TYPE",
+  "anonymous": $ANONYMOUS,
+  "project_context": "$PROJECT_CONTEXT",
+  "saved_at": "$TIMESTAMP"
+}
+EOF
+```
+
+Inform the user:
+```
+[Saved] Feedback saved locally to ~/.omcodex/feedback/{timestamp}.json
+Submit manually when connectivity is available:
+  - GitHub Issues: https://github.com/baekenough/oh-my-customcodex/issues/new
+  - Or run /omcodex:feedback again when gh is available
+```
+
+### Category-to-Label Mapping
+
+| Category | GitHub Label |
+|----------|--------------|
+| bug | bug |
+| feature | enhancement |
+| improvement | enhancement |
+| question | question |
+| (auto-detect fails) | (none) |
+
+## Notes
+
+- Route A creates a visible GitHub issue attributed to the user's gh account
+- When `--anonymous` is used, the title is prefixed with `[Anonymous Feedback]` and the `anonymous` label is added
+- Fallback ensures no feedback is silently lost even in offline environments
+- `disable-model-invocation: true` ensures this skill only runs when explicitly invoked by the user
+- Target repo is hardcoded to `baekenough/oh-my-customcodex` — feedback is always about the child package itself

package/templates/.claude/skills/omcodex-improve-report/SKILL.md

@@ -0,0 +1,65 @@
+---
+name: omcodex:improve-report
+description: Read-only report of improvement suggestions from eval-core analysis engine
+scope: harness
+user-invocable: true
+---
+
+# Improve Report
+
+Display improvement suggestions from eval-core analysis engine as read-only report. No file modifications, no GitHub mutations.
+
+## Purpose
+
+Surface actionable improvement suggestions gathered by the eval-core analysis engine. Reads from eval-core's local database and renders insights as structured markdown. Useful for understanding routing quality, skill effectiveness, and agent usage patterns.
+
+## Usage
+
+```
+/omcodex:improve-report
+```
+
+## Workflow
+
+### Step 1: Check eval-core availability
+
+```bash
+command -v eval-core 2>/dev/null || ls node_modules/.bin/eval-core 2>/dev/null
+```
+
+If not found → skip to Step 4 (fallback).
+
+### Step 2: Run analysis
+
+```bash
+eval-core analyze --format markdown
+```
+
+### Step 3: Data sufficiency check
+
+Count sessions recorded from the output metadata:
+
+| Sessions recorded | Confidence | Behavior |
+|-------------------|------------|----------|
+| < 5 | `[confidence: low]` | Show results with message: "최소 5세션 이상의 데이터가 필요합니다. 현재 데이터로도 초기 분석은 가능하지만, 정확도는 제한적입니다." |
+| 5–20 | `[confidence: medium]` | Show results with medium confidence note |
+| > 20 | `[confidence: high]` | Show results normally |
+
+### Step 4: Display
+
+If output contains data → display as structured markdown with confidence annotation prepended.
+
+If command not found OR output is empty:
+
+```
+[omcodex:improve-report] 데이터 없음
+eval-core가 설치되어 있지 않거나 아직 충분한 세션 데이터가 수집되지 않았습니다.
+eval-core를 설치하거나 더 많은 세션을 진행한 후 다시 실행하세요.
+```
+
+## Notes
+
+- **Read-only**: No file modifications, no GitHub issue creation, no external mutations
+- **R010 compliant**: Only executes Bash for read-only analysis — no write operations delegated
+- **Graceful fallback**: Falls back silently if eval-core or its DB doesn't exist
+- **Confidence-aware**: Reports data sufficiency alongside results per R011 conventions

package/templates/.claude/skills/omcodex-loop/SKILL.md

@@ -0,0 +1,45 @@
+---
+name: omcodex:loop
+description: Prevent session idle during background agent work via SubagentStop prompt hook auto-continuation
+scope: core
+version: 1.0.0
+user-invocable: true
+---
+
+# /omcodex:loop — Session Auto-Continuation
+
+## Overview
+
+Prevents session idle when background subagents complete by using a `SubagentStop` prompt hook that nudges the orchestrator to check for pending workflow steps.
+
+## How It Works
+
+1. When a background subagent completes, the runtime fires the `SubagentStop` event
+2. The prompt hook injects a message asking the orchestrator to check for pending steps
+3. If pending steps exist, the orchestrator proceeds automatically
+4. If no pending steps, it reports results and waits for user input
+
+## Safety Limits
+
+- **3 consecutive auto-continues max**: After 3 automatic progressions without user interaction, the system pauses and asks the user before proceeding
+- **stuck-detector integration**: If the same action repeats 3+ times, stuck-detector intervenes
+- **cost-cap-advisor**: Cost monitoring continues during auto-continuation
+
+## Configuration
+
+The hook is configured in `.codex/hooks/hooks.json` under `SubagentStop`. It works alongside the existing `task-outcome-recorder.sh` command hook.
+
+## Limitations
+
+- **Platform constraint**: The turn-based runtime model means the prompt hook only fires when a subagent completes — it cannot wake the model from true idle state
+- **Foreground agents preferred**: For guaranteed continuation, use foreground parallel agents (R009) instead of background agents
+- **PoC status**: This is an experimental feature. If the prompt hook doesn't reliably trigger in all scenarios, fall back to foreground agent patterns
+
+## Usage
+
+```bash
+/omcodex:loop          # Show current auto-continuation status
+/omcodex:loop status   # Same as above
+```
+
+The feature is active by default via hooks.json. No explicit activation needed.