oh-my-customcode 0.101.0 → 0.103.0

package/README.md

@@ -13,7 +13,7 @@
 
 **[한국어 문서 (Korean)](./README_ko.md)**
 
-48 agents. 109 skills. 22 rules. One command.
+48 agents. 112 skills. 22 rules. One command.
 
 ```bash
 npm install -g oh-my-customcode && cd your-project && omcustom init
@@ -132,7 +132,7 @@ Each agent declares its tools, model, memory scope, and limitations in YAML fron
 
 ---
 
-### Skills (109)
+### Skills (112)
 
 | Category | Count | Includes |
 |----------|-------|----------|
@@ -222,7 +222,7 @@ Key rules: R010 (orchestrator never writes files), R009 (parallel execution mand
 
 ---
 
-### Guides (39)
+### Guides (40)
 
 Reference documentation covering best practices, architecture decisions, and integration patterns. Located in `guides/` at project root, covering topics from agent design to CI/CD to observability.
 
@@ -272,14 +272,14 @@ your-project/
 ├── CLAUDE.md                   # Entry point
 ├── .claude/
 │   ├── agents/                 # 48 agent definitions
-│   ├── skills/                 # 109 skill modules
+│   ├── skills/                 # 112 skill modules
 │   ├── rules/                  # 22 governance rules (R000-R021)
 │   ├── hooks/                  # 15 lifecycle hook scripts
 │   ├── schemas/                # Tool input validation schemas
 │   ├── specs/                  # Extracted canonical specs
 │   ├── contexts/               # 4 shared context files
 │   └── ontology/               # Knowledge graph for RAG
-└── guides/                     # 39 reference documents
+└── guides/                     # 40 reference documents
 ```
 
 ---

package/dist/cli/index.js

@@ -2334,7 +2334,7 @@ var init_package = __esm(() => {
     workspaces: [
       "packages/*"
     ],
-    version: "0.101.0",
+    version: "0.103.0",
     description: "Batteries-included agent harness for Claude Code",
     type: "module",
     bin: {

package/dist/index.js

@@ -2014,7 +2014,7 @@ var package_default = {
   workspaces: [
     "packages/*"
   ],
-  version: "0.101.0",
+  version: "0.103.0",
   description: "Batteries-included agent harness for Claude Code",
   type: "module",
   bin: {

package/package.json

@@ -3,7 +3,7 @@
   "workspaces": [
     "packages/*"
   ],
-  "version": "0.101.0",
+  "version": "0.103.0",
   "description": "Batteries-included agent harness for Claude Code",
   "type": "module",
   "bin": {

package/templates/.claude/agents/arch-speckit-agent.md

@@ -26,7 +26,7 @@ You are a Spec-Driven Development agent that transforms requirements into execut
 External agent from https://github.com/github/spec-kit
 - **Version**: latest
 - **Update**: `uv tool upgrade specify-cli --from git+https://github.com/github/spec-kit.git`
-- **Prerequisites**: Python 3.11+, uv, Git
+- **Prerequisites**: Python 3.14+, uv, Git
 
 ## Commands
 

package/templates/.claude/agents/be-springboot-expert.md

@@ -1,6 +1,6 @@
 ---
 name: be-springboot-expert
-description: Expert Spring Boot 3.5.x developer for enterprise-grade Java 21 applications. Use for Spring Boot projects, Java/Kotlin enterprise apps, RESTful APIs, microservices architecture, Spring Data, security patterns, virtual threads, and GraalVM native images.
+description: Expert Spring Boot 4.0 developer for enterprise-grade Java 25 LTS applications. Use for Spring Boot projects, Java/Kotlin enterprise apps, RESTful APIs, microservices architecture, Spring Data, security patterns, virtual threads, and GraalVM native images.
 model: sonnet
 domain: backend
 memory: project
@@ -17,7 +17,7 @@ tools:
 permissionMode: bypassPermissions
 ---
 
-You are an expert Spring Boot developer for enterprise-grade Java/Kotlin applications. Focused on Spring Boot 3.5.x with Java 21.
+You are an expert Spring Boot developer for enterprise-grade Java/Kotlin applications. Focused on Spring Boot 4.0 (GA: 2025-11) with Java 25 LTS.
 
 ## Capabilities
 
@@ -31,7 +31,7 @@ You are an expert Spring Boot developer for enterprise-grade Java/Kotlin applica
 ## Reference Documentation
 
 - https://docs.spring.io/spring-boot/reference/index.html
-- https://github.com/spring-projects/spring-boot/wiki/Spring-Boot-3.5-Release-Notes
+- https://github.com/spring-projects/spring-boot/wiki/Spring-Boot-4.0-Release-Notes
 
 ## Skills
 

package/templates/.claude/agents/de-airflow-expert.md

@@ -17,7 +17,7 @@ tools:
 permissionMode: bypassPermissions
 ---
 
-You are an expert Apache Airflow developer for production-ready DAGs following official best practices, targeting **Airflow 3.1.8**.
+You are an expert Apache Airflow developer for production-ready DAGs following official best practices, targeting **Airflow 3.2.0**.
 
 ## Capabilities
 
@@ -41,7 +41,7 @@ You are an expert Apache Airflow developer for production-ready DAGs following o
 
 ## Skills
 
-Apply **airflow-best-practices** for core Airflow 3.1.8 guidelines.
+Apply **airflow-best-practices** for core Airflow 3.2.0 guidelines.
 
 ## Reference Guides
 

package/templates/.claude/agents/de-dbt-expert.md

@@ -21,6 +21,7 @@ You are an expert dbt developer for analytics engineering, SQL modeling, and dat
 
 ## Capabilities
 
+- dbt-core 1.11+ with Fusion engine (faster parallel execution)
 - Project structure: staging (stg_), intermediate (int_), marts (fct_, dim_)
 - Materializations (view, ephemeral, table, incremental)
 - Schema tests (unique, not_null, relationships, accepted_values)

package/templates/.claude/agents/de-kafka-expert.md

@@ -17,7 +17,7 @@ tools:
 permissionMode: bypassPermissions
 ---
 
-You are an expert Apache Kafka developer for event streaming architectures with high throughput and reliability.
+You are an expert Apache Kafka 4.x developer for event streaming architectures with high throughput and reliability. Kafka 4.x uses KRaft (Kafka Raft) as the default metadata management — ZooKeeper has been fully removed.
 
 ## Capabilities
 
@@ -26,6 +26,7 @@ You are an expert Apache Kafka developer for event streaming architectures with
 - Consumer applications with proper offset management
 - Kafka Streams and Connect pipelines
 - Schema Registry with Avro/Protobuf serialization
+- KRaft cluster deployment and configuration (ZooKeeper removed in Kafka 4.x)
 - Cluster performance and monitoring
 - Event-driven architectures and CQRS patterns
 

package/templates/.claude/agents/de-spark-expert.md

@@ -17,17 +17,19 @@ tools:
 permissionMode: bypassPermissions
 ---
 
-You are an expert Apache Spark developer for performant distributed data processing using PySpark and Scala.
+You are an expert Apache Spark 4.x developer for performant distributed data processing using PySpark and Scala. Spark 4.0 introduces Spark Connect as a decoupled client-server protocol and enables ANSI mode by default.
 
 ## Capabilities
 
 - DataFrame and Dataset APIs, Spark SQL
+- Spark Connect for decoupled client-server communication (Spark 4.x)
 - Broadcast joins and hint-based optimization
 - Partitioning and bucketing strategies
 - Structured Streaming applications
 - Resource management (executor/driver memory, dynamic allocation)
 - Storage format optimization (Parquet, ORC, Delta, Iceberg)
 - Job profiling via Spark UI
+- ANSI mode compliance (default in Spark 4.x — stricter SQL semantics)
 
 ## Skills
 

package/templates/.claude/agents/fe-flutter-agent.md

@@ -22,8 +22,8 @@ You are an expert Flutter developer following official documentation and Dart be
 ## Capabilities
 
 - Widget composition (StatelessWidget, StatefulWidget, InheritedWidget)
-- State management: Riverpod 3.0 (default), BLoC 9.0 (enterprise)
-- Dart 3.x null safety, sealed classes, pattern matching, records
+- State management: Riverpod 3.3 (default), BLoC 9.1 (enterprise)
+- Dart 3.11 null safety, sealed classes, pattern matching, records
 - go_router declarative navigation with deep linking
 - freezed immutable data models with code generation
 - Performance: const constructors, RepaintBoundary, Isolate
@@ -32,7 +32,7 @@ You are an expert Flutter developer following official documentation and Dart be
 
 ## Default Stack
 
-- **State**: Riverpod 3.0
+- **State**: Riverpod 3.3
 - **Navigation**: go_router
 - **Models**: freezed + json_serializable
 - **HTTP**: dio

package/templates/.claude/agents/lang-java-expert.md

@@ -0,0 +1,42 @@
+---
+name: lang-java-expert
+description: Expert Java 25 LTS developer for modern Java with Virtual Threads, Pattern Matching, Record Patterns, and Sequenced Collections. Use for Java files (*.java, pom.xml, build.gradle), Java-related keywords, high-concurrency systems, data-centric APIs with Records, and migrating from older Java versions.
+model: sonnet
+domain: backend
+memory: project
+effort: high
+skills:
+  - java-best-practices
+tools:
+  - Read
+  - Write
+  - Edit
+  - Grep
+  - Glob
+  - Bash
+permissionMode: bypassPermissions
+---
+
+You are an expert Java 25 LTS developer for modern Java features including Virtual Threads, Pattern Matching, Record Patterns, and Sequenced Collections.
+
+## Capabilities
+
+- Modern Java 25 LTS leveraging latest features
+- Virtual Threads (JEP 444) for scalable concurrency
+- Pattern Matching for switch and instanceof
+- Record Patterns (JEP 440), Sequenced Collections (JEP 431)
+- Google Java Style Guide compliance
+- Legacy Java migration to Java 25 LTS
+
+## Reference Documentation
+
+- https://docs.oracle.com/en/java/javase/25/
+- https://google.github.io/styleguide/javaguide.html
+
+## Skills
+
+Apply **java-best-practices** for core Java guidelines.
+
+## Reference Guides
+
+Consult `guides/java/` for reference documentation.

package/templates/.claude/agents/slack-cli-expert.md

@@ -14,7 +14,7 @@ tools:
 permissionMode: bypassPermissions
 ---
 
-You are an expert Slack CLI developer specialized in building, deploying, and managing Slack apps using the official Slack CLI and the Slack Platform (https://docs.slack.dev/tools/slack-cli/).
+You are an expert Slack CLI developer specialized in building, deploying, and managing Slack apps using the official Slack CLI v4.0 and the Slack Platform (https://docs.slack.dev/tools/slack-cli/). Slack CLI v4.0 adds first-class Slack Agent development support.
 
 ## Capabilities
 

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

@@ -475,6 +475,16 @@
           }
         ],
         "description": "Advisory reminder to sync skill counts in 6 locations when a SKILL.md is created/modified (R021)"
+      },
+      {
+        "matcher": "mcp_tool_name matches \"mcp__playwright__.*\" || mcp_tool_name matches \"mcp__claude-in-chrome__.*\"",
+        "hooks": [
+          {
+            "type": "command",
+            "command": "bash .claude/hooks/scripts/playwright-compress.sh"
+          }
+        ],
+        "description": "Layer 4: Compress Playwright/Chrome MCP output via Haiku summarization"
       }
     ],
     "Stop": [

package/templates/.claude/hooks/scripts/playwright-compress.sh

@@ -0,0 +1,50 @@
+#!/bin/bash
+# Layer 4: Playwright/Chrome MCP Output Intelligence Compression
+# Reduces MCP tool output by 94-96% using Haiku summarization
+# Preserves ref= values for interactive flow continuity
+# Source: adapted from treesoop/claude-native-plugin (MIT)
+
+set -euo pipefail
+
+input=$(cat)
+tool_output=$(echo "$input" | jq -r '.tool_output // ""')
+
+# Skip if output is small (< 3000 chars)
+output_len=${#tool_output}
+if [ "$output_len" -lt 3000 ]; then
+  echo "$input"
+  exit 0
+fi
+
+# Extract ref= values to preserve
+refs=$(echo "$tool_output" | grep -oE 'ref="[^"]*"' | sort -u || true)
+
+# Summarize using Haiku via subscription auth
+summary=$(echo "$tool_output" | claude -p --model haiku "Summarize this browser page content concisely. Preserve ALL ref= attribute values exactly as they appear. Focus on: page structure, interactive elements with their ref values, visible text content, and any error messages." 2>/dev/null) || {
+  # Fallback: return original on failure
+  echo "$input"
+  exit 0
+}
+
+# Verify ref= preservation
+if [ -n "$refs" ]; then
+  missing_refs=""
+  while IFS= read -r ref; do
+    if ! echo "$summary" | grep -qF "$ref"; then
+      missing_refs="$missing_refs $ref"
+    fi
+  done <<< "$refs"
+
+  # Append missing refs if any
+  if [ -n "$missing_refs" ]; then
+    summary="$summary
+
+[Preserved refs]:$missing_refs"
+  fi
+fi
+
+# Return compressed output
+compressed_len=${#summary}
+savings=$(( (output_len - compressed_len) * 100 / output_len ))
+echo "$input" | jq --arg summary "$summary" --arg savings "${savings}% reduced (${output_len}→${compressed_len} chars)" \
+  '.tool_output = $summary | .["updatedMCPToolOutput"] = $summary'

package/templates/.claude/ontology/agents.yaml

@@ -6,7 +6,7 @@ generated_at: "2026-02-12T00:00:00Z"
 # Agent class hierarchy
 classes:
   LanguageExpert:
-    agents: [lang-golang-expert, lang-python-expert, lang-rust-expert, lang-kotlin-expert, lang-typescript-expert, lang-java21-expert]
+    agents: [lang-golang-expert, lang-python-expert, lang-rust-expert, lang-kotlin-expert, lang-typescript-expert, lang-java-expert]
     description: "Programming language specialists"
   BackendExpert:
     agents: [be-fastapi-expert, be-springboot-expert, be-go-backend-expert, be-express-expert, be-nestjs-expert]
@@ -101,7 +101,7 @@ agents:
     keywords: [typescript, type-safety, generics, conditional-types, mapped-types, nodejs]
     file_patterns: ["*.ts", "*.tsx", "tsconfig.json", "package.json"]
 
-  lang-java21-expert:
+  lang-java-expert:
     class: LanguageExpert
     description: "Expert Java 21 developer for modern Java with Virtual Threads, Pattern Matching, Record Patterns, and Sequenced Collections. Use for Java files (*.java, pom.xml, build.gradle), Java-related keywords, high-concurrency systems, data-centric APIs with Records, and migrating from older Java versions."
     model: sonnet

package/templates/.claude/ontology/skills.yaml

@@ -147,7 +147,7 @@ skills:
     keywords: [routing, development, code-review, implementation, refactoring]
     routes_to: [
       lang-golang-expert, lang-python-expert, lang-rust-expert, lang-kotlin-expert,
-      lang-typescript-expert, lang-java21-expert, fe-vercel-agent, fe-vuejs-agent,
+      lang-typescript-expert, lang-java-expert, fe-vercel-agent, fe-vuejs-agent,
       fe-svelte-agent, fe-flutter-agent, be-fastapi-expert, be-springboot-expert,
       be-go-backend-expert, be-nestjs-expert, be-express-expert, tool-npm-expert,
       tool-optimizer, tool-bun-expert, db-supabase-expert, db-postgres-expert,

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

@@ -227,6 +227,35 @@ When the user explicitly signals full-delegation intent, the orchestrator operat
 Autonomous mode and `/structured-dev-cycle` (stage-blocker) are mutually exclusive.
 -->
 
+## Universal bypassPermissions
+
+**ALL Agent tool calls MUST include `mode: "bypassPermissions"`.**
+
+The Agent tool defaults to `mode: "acceptEdits"`, which overrides agent frontmatter `permissionMode` and causes permission prompts during unattended execution. This is a CC platform behavior, not a configuration error.
+
+| Aspect | Detail |
+|--------|--------|
+| Scope | Every Agent tool call, without exception |
+| Why | CC's Agent tool `mode` default (`acceptEdits`) overrides frontmatter |
+| History | #926 (v0.99.1), #947 (v0.100.1), #955 (v0.103.0) — recurring issue |
+| Enforcement | Prompt-based (R021); all 23 agent-spawning skills include instruction |
+
+### Self-Check
+
+Before spawning any agent:
+1. Does the Agent tool call include `mode: "bypassPermissions"`? → YES: proceed → NO: add it
+2. Is this a new skill that spawns agents? → Add Permission Mode section
+
+### Common Violation
+
+```
+❌ WRONG: Agent tool call without mode parameter
+   Agent(subagent_type: "lang-golang-expert", prompt: "...")
+
+✓ CORRECT: Always include mode
+   Agent(subagent_type: "lang-golang-expert", mode: "bypassPermissions", prompt: "...")
+```
+
 ## Session Continuity
 
 After restart/compaction: re-read CLAUDE.md, all delegation rules still apply. Never write code directly from orchestrator.

package/templates/.claude/skills/adaptive-harness/SKILL.md

@@ -102,7 +102,7 @@ Check for language manifest files and framework indicators:
 | `svelte.config.*`, `*.svelte` | Svelte | fe-svelte-agent |
 | `pubspec.yaml`, `*.dart` | Flutter | fe-flutter-agent |
 | `*.kt`, `build.gradle.kts` | Kotlin | lang-kotlin-expert |
-| `*.java`, `pom.xml` | Java | lang-java21-expert |
+| `*.java`, `pom.xml` | Java | lang-java-expert |
 | `spring-boot` in deps | Spring Boot | be-springboot-expert |
 | `express` in package.json deps | Express | be-express-expert |
 | `@nestjs` in package.json deps | NestJS | be-nestjs-expert |

package/templates/.claude/skills/agora/SKILL.md

@@ -192,3 +192,5 @@ agora:
 | Accepting first BUILD | May miss edge cases | Loop until ALL agree |
 | Ignoring split decisions | Unresolved disagreements fester | Resolve every split with evidence |
 | Push for consensus too fast | Premature agreement | Let reviewers challenge freely |
+
+When spawning agents via the Agent tool during this skill's execution, always pass `mode: "bypassPermissions"`. The Agent tool default (`acceptEdits`) overrides agent frontmatter `permissionMode`, causing permission prompts during unattended execution.

package/templates/.claude/skills/airflow-best-practices/SKILL.md

@@ -5,7 +5,7 @@ scope: core
 user-invocable: false
 ---
 
-# Apache Airflow Best Practices (3.1.8)
+# Apache Airflow Best Practices (3.2.0)
 
 ## DAG Authoring
 
@@ -89,6 +89,6 @@ user-invocable: false
 - **New UI**: React-based web interface (replaces Flask-based UI)
 
 ## References
-- [Airflow 3.1.8 Best Practices](https://airflow.apache.org/docs/apache-airflow/3.1.8/best-practices.html)
-- [Airflow SDK (Task SDK)](https://airflow.apache.org/docs/apache-airflow/3.1.8/authoring-and-scheduling/index.html)
-- [Migration Guide 2.x → 3.x](https://airflow.apache.org/docs/apache-airflow/3.1.8/migration-guide.html)
+- [Airflow 3.2.0 Best Practices](https://airflow.apache.org/docs/apache-airflow/3.2.0/best-practices.html)
+- [Airflow SDK (Task SDK)](https://airflow.apache.org/docs/apache-airflow/3.2.0/authoring-and-scheduling/index.html)
+- [Migration Guide 2.x → 3.x](https://airflow.apache.org/docs/apache-airflow/3.2.0/migration-guide.html)

package/templates/.claude/skills/analysis/SKILL.md

@@ -76,7 +76,7 @@ Detect tech stack by checking indicator files and dependency manifests.
 | Go Backend | go.mod + cmd/ or internal/ dirs | be-go-backend-expert | go-backend-best-practices |
 | Rust | Cargo.toml, *.rs | lang-rust-expert | rust-best-practices |
 | Kotlin | *.kt, build.gradle.kts | lang-kotlin-expert | kotlin-best-practices |
-| Java | *.java, pom.xml | lang-java21-expert | - |
+| Java | *.java, pom.xml | lang-java-expert | - |
 | Spring Boot | spring-boot in deps | be-springboot-expert | springboot-best-practices |
 | Express.js | "express" in deps | be-express-expert | - |
 | NestJS | "@nestjs" in deps | be-nestjs-expert | - |

package/templates/.claude/skills/dbt-best-practices/SKILL.md

@@ -51,5 +51,10 @@ user-invocable: false
 - Use `doc` blocks for reusable text
 - Generate and host dbt docs
 
+## Version
+
+- **dbt-core**: 1.11+ (includes Fusion engine for faster parallel model execution)
+- Fusion engine is enabled by default in dbt-core 1.11 — no configuration required
+
 ## References
 - [dbt Best Practices](https://docs.getdbt.com/guides/best-practices)

package/templates/.claude/skills/design-shotgun/SKILL.md

@@ -0,0 +1,104 @@
+---
+name: design-shotgun
+description: Generate 4-6 parallel design mockups for rapid visual comparison — adapted from gstack /design-shotgun pattern
+scope: core
+version: 1.0.0
+user-invocable: true
+argument-hint: "<component/page description>"
+effort: high
+---
+
+# Design Shotgun — Parallel Mockup Generation
+
+## Purpose
+
+Generates 4-6 independent design variations simultaneously, then presents them side-by-side for comparison. Prevents premature convergence on a single design direction.
+
+## Usage
+
+```
+/design-shotgun "landing page hero section"
+/design-shotgun "dashboard settings panel"
+/design-shotgun "mobile navigation menu"
+```
+
+## Workflow
+
+### Phase 1: Brief Analysis
+
+Parse the design brief to extract:
+- Component type (page, section, widget, modal)
+- Constraints (brand colors, existing design system, accessibility requirements)
+- Target platform (web, mobile, responsive)
+
+### Phase 2: Parallel Generation (R009)
+
+Spawn 4 parallel agents, each generating a distinct design approach:
+
+| Agent | Style Direction | Focus |
+|-------|----------------|-------|
+| 1 | Minimal | Maximum whitespace, essential elements only |
+| 2 | Data-dense | Information-rich, compact layout |
+| 3 | Visual | Hero imagery, bold typography, emotional |
+| 4 | Conventional | Industry-standard patterns, familiar UX |
+
+Each agent generates:
+- HTML mockup (self-contained, inline CSS)
+- Design rationale (2-3 sentences)
+- Accessibility notes
+
+### Phase 3: Comparison Board
+
+Present all mockups with side-by-side comparison:
+
+```markdown
+## Design Shotgun Results: {component}
+
+### Variation 1: Minimal
+Rationale: {why}
+[HTML mockup code]
+
+### Variation 2: Data-dense
+Rationale: {why}
+[HTML mockup code]
+
+### Variation 3: Visual
+Rationale: {why}
+[HTML mockup code]
+
+### Variation 4: Conventional
+Rationale: {why}
+[HTML mockup code]
+
+## Comparison
+| Criteria | V1 | V2 | V3 | V4 |
+|----------|----|----|----|----|
+| Readability | ★★★ | ★★ | ★★ | ★★★ |
+| Visual impact | ★ | ★★ | ★★★ | ★★ |
+| Information density | ★ | ★★★ | ★★ | ★★ |
+| Accessibility | ★★★ | ★★ | ★★ | ★★★ |
+```
+
+### Phase 4: Selection & Refinement
+
+User selects preferred variation(s). Selected design can be:
+- Refined with follow-up iterations
+- Combined with elements from other variations
+- Handed to fe-design-expert for production implementation
+
+## Integration
+
+| Component | Role |
+|-----------|------|
+| R009 | 4 parallel agents for mockup generation |
+| impeccable-design | Quality baseline for generated mockups |
+| fe-design-expert | Production refinement after selection |
+| web-design-guidelines | Accessibility and UX compliance |
+
+## Permission Mode
+
+When spawning agents via the Agent tool during this skill's execution, always pass `mode: "bypassPermissions"`. The Agent tool default (`acceptEdits`) overrides agent frontmatter `permissionMode`, causing permission prompts during unattended execution.
+
+## Source
+
+Adapted from [garrytan/gstack](https://github.com/garrytan/gstack) /design-shotgun pattern.

package/templates/.claude/skills/dev-lead-routing/SKILL.md

@@ -12,7 +12,7 @@ context: fork
 
 | Type | Agents |
 |------|--------|
-| Language | lang-golang-expert, lang-python-expert, lang-rust-expert, lang-kotlin-expert, lang-typescript-expert, lang-java21-expert |
+| Language | lang-golang-expert, lang-python-expert, lang-rust-expert, lang-kotlin-expert, lang-typescript-expert, lang-java-expert |
 | Frontend | fe-vercel-agent, fe-vuejs-agent, fe-svelte-agent, fe-flutter-agent, fe-design-expert |
 | Backend | be-fastapi-expert, be-springboot-expert, be-go-backend-expert, be-nestjs-expert, be-express-expert, be-django-expert |
 | Tooling | tool-npm-expert, tool-optimizer, tool-bun-expert |
@@ -31,7 +31,7 @@ context: fork
 | `.rs` | lang-rust-expert |
 | `.kt`, `.kts` | lang-kotlin-expert |
 | `.ts`, `.tsx` | lang-typescript-expert |
-| `.java` | lang-java21-expert |
+| `.java` | lang-java-expert |
 | `.js/.jsx` (React) | fe-vercel-agent |
 | `.vue` | fe-vuejs-agent |
 | `.svelte` | fe-svelte-agent |
@@ -52,7 +52,7 @@ context: fork
 | rust | lang-rust-expert |
 | kotlin | lang-kotlin-expert |
 | typescript, ts | lang-typescript-expert |
-| java | lang-java21-expert |
+| java | lang-java-expert |
 | react, next.js, vercel | fe-vercel-agent |
 | vue | fe-vuejs-agent |
 | svelte | fe-svelte-agent |

package/templates/.claude/skills/django-best-practices/SKILL.md

@@ -7,7 +7,7 @@ user-invocable: false
 
 ## Purpose
 
-Apply Django patterns for building production-ready, secure, and maintainable Python web applications.
+Apply Django 6.0.4 patterns for building production-ready, secure, and maintainable Python web applications.
 
 ## Rules
 

package/templates/.claude/skills/evaluator-optimizer/SKILL.md

@@ -419,3 +419,5 @@ Integration: Works with [impeccable-design](/skills/impeccable-design) skill for
 ### Harness Eval Preset
 
 The `harness-eval` skill provides a structured 15-task SE benchmark rubric that can be used as a preset for the evaluator-optimizer pipeline. When invoked via `/omcustom:harness-eval`, the harness rubric dimensions (Test Coverage 30%, Architecture 25%, Error Handling 25%, Extensibility 20%) are loaded as the sprint contract criteria.
+
+When spawning agents via the Agent tool during this skill's execution, always pass `mode: "bypassPermissions"`. The Agent tool default (`acceptEdits`) overrides agent frontmatter `permissionMode`, causing permission prompts during unattended execution.

package/templates/.claude/skills/flutter-best-practices/SKILL.md

@@ -81,8 +81,8 @@ Reference: guides/flutter/fundamentals.md
 
 ```yaml
 default_choice:
-  new_projects: "Riverpod 3.0"
-  enterprise: "BLoC 9.0"
+  new_projects: "Riverpod 3.3"
+  enterprise: "BLoC 9.1"
   simple_prototypes: "setState or Provider"
   avoid: "GetX (maintenance crisis, runtime crashes)"
 
@@ -285,7 +285,7 @@ Reference: guides/flutter/architecture.md
 ## Default Stack
 
 ```yaml
-state_management: Riverpod 3.0
+state_management: Riverpod 3.3
 navigation: go_router
 models: freezed + json_serializable
 di: Riverpod (built-in)
@@ -298,7 +298,7 @@ structure: Official MVVM (lib/{ui,data}/)
 ## Enterprise Stack
 
 ```yaml
-state_management: BLoC 9.0 + Cubit
+state_management: BLoC 9.1 + Cubit
 navigation: go_router or auto_route
 models: freezed + json_serializable
 di: GetIt + injectable (or Riverpod)
@@ -314,7 +314,7 @@ When writing or reviewing Flutter/Dart code:
 1. **Always** use const constructors for static widgets
 2. **Always** return Result<T> from repositories, never throw
 3. **Always** use flutter_secure_storage for sensitive data
-4. **Prefer** Riverpod 3.0 for new projects, BLoC for enterprise
+4. **Prefer** Riverpod 3.3 for new projects, BLoC for enterprise
 5. **Prefer** StatelessWidget classes over helper functions
 6. **Prefer** sealed classes for state/result types (exhaustive matching)
 7. **Use** freezed for all data model classes

package/templates/.claude/skills/intent-detection/SKILL.md

@@ -306,3 +306,5 @@ Check Codex CLI availability
 ├── Method: codex-exec (xhigh) | WebFetch fallback
 └── Reason: {explanation}
 ```
+
+When spawning agents via the Agent tool during this skill's execution, always pass `mode: "bypassPermissions"`. The Agent tool default (`acceptEdits`) overrides agent frontmatter `permissionMode`, causing permission prompts during unattended execution.

package/templates/.claude/skills/intent-detection/patterns/agent-triggers.yaml

@@ -69,10 +69,10 @@ agents:
     supported_actions: [review, fix, refactor, explain]
     base_confidence: 40
 
-  lang-java21-expert:
+  lang-java-expert:
     keywords:
-      korean: [자바, 자바21]
-      english: [java, java21, jdk21, jdk]
+      korean: [자바, 자바25]
+      english: [java, java25, jdk25, jdk]
     file_patterns: ["*.java", "pom.xml", "build.gradle"]
     supported_actions: [review, create, fix, refactor]
     base_confidence: 35