agentic-code 0.6.0 → 0.6.2

package/.agents/skills/ai-development-guide/SKILL.md

@@ -1,8 +1,8 @@
 ---
 name: ai-development-guide
-description: |
-  Technical anti-pattern detection, fail-fast design principles, debugging techniques, and implementation completeness guidelines.
-  Use when: implementing features, reviewing code quality, debugging errors, refactoring code, or detecting code smells.
+description: >
+  Detects code smells, anti-patterns, and debugging issues. Guides fail-fast design and implementation completeness.
+  Use when: fixing bugs, reviewing code quality, refactoring, or detecting code smells.
 ---
 
 # AI Developer Guide

package/.agents/skills/coding-rules/SKILL.md

@@ -1,8 +1,8 @@
 ---
 name: coding-rules
-description: |
-  Coding principles for maintainability, readability, and quality.
-  Use when: implementing features, refactoring code, reviewing code quality, or establishing coding standards.
+description: >
+  Applies coding standards for clean, maintainable code. Covers functions, error handling, and refactoring techniques.
+  Use when: writing functions, handling errors, refactoring, or reviewing code style.
 ---
 
 # Development Rules

package/.agents/skills/documentation-criteria/SKILL.md

@@ -1,8 +1,8 @@
 ---
 name: documentation-criteria
-description: |
-  Documentation creation criteria including PRD, ADR, Design Doc, and Work Plan requirements with templates.
-  Use when: creating technical documents, determining which documents are required, reviewing document quality, or planning feature implementation.
+description: >
+  Guides PRD, ADR, Design Doc, and Work Plan creation with templates and decision matrix.
+  Use when: planning features, writing specs, creating technical documents, or deciding which docs are needed.
 ---
 
 # Documentation Creation Criteria

package/.agents/skills/implementation-approach/SKILL.md

@@ -1,8 +1,8 @@
 ---
 name: implementation-approach
-description: |
-  Implementation strategy selection framework with meta-cognitive approach, verification levels, and integration point definitions.
-  Use when: planning implementation strategy, selecting development approach (vertical/horizontal/hybrid), or defining verification criteria.
+description: >
+  Selects implementation strategy (vertical slice, horizontal slice, or hybrid) with risk assessment.
+  Use when: planning how to build a feature, deciding development approach, or defining verification levels.
 ---
 
 # Implementation Strategy Selection Framework (Meta-cognitive Approach)

package/.agents/skills/integration-e2e-testing/SKILL.md

@@ -1,8 +1,8 @@
 ---
 name: integration-e2e-testing
-description: |
-  Integration and E2E test design principles, behavior verification rules, mock boundaries, and review criteria.
-  Use when: designing integration tests, E2E tests, writing test skeletons, or reviewing test implementation quality.
+description: >
+  Designs integration and E2E tests with mock boundaries and behavior verification rules.
+  Use when: writing E2E tests, integration tests, test skeletons, or reviewing test quality.
 ---
 
 # Integration Test & E2E Test Design/Implementation Rules

package/.agents/skills/metacognition/SKILL.md

@@ -1,8 +1,8 @@
 ---
 name: metacognition
-description: |
-  Self-assessment checkpoints and protocol for AI agents.
-  Use when: task type changes, after completing tasks, encountering errors, before starting new features, or switching between phases.
+description: >
+  Validates approach and checks assumptions before and after tasks. Provides self-assessment protocol.
+  Use when: starting new work, encountering errors, completing tasks, or switching between phases.
 ---
 
 # Metacognition Protocol

package/.agents/skills/testing/SKILL.md

@@ -1,8 +1,8 @@
 ---
 name: testing
-description: |
-  Testing principles including TDD process, test quality criteria, coverage standards, and mock usage guidelines.
-  Use when: writing tests, designing test strategies, reviewing test quality, or establishing testing standards.
+description: >
+  Applies TDD process, test quality criteria, and mock usage guidelines. Covers unit, integration, and E2E tests.
+  Use when: writing unit tests, using mocks, reviewing test quality, or ensuring test coverage.
 ---
 
 # Testing Rules

package/.agents/skills/testing-strategy/SKILL.md

@@ -1,8 +1,8 @@
 ---
 name: testing-strategy
-description: |
-  ROI-based test selection framework with calculation formula, critical user journey definitions, and test prioritization guidelines.
-  Use when: designing test strategy, selecting which tests to write, calculating test ROI, or prioritizing test coverage.
+description: >
+  Prioritizes which tests to write based on ROI calculation and critical user journeys.
+  Use when: deciding test coverage, selecting test types, or budgeting test effort.
 ---
 
 # Test Strategy: ROI-Based Selection

package/AGENTS.md

@@ -20,7 +20,12 @@
 
 ## Task Analysis - Required First Step
 
-**ALWAYS start here for any user request:**
+**Early Exit Check (BEFORE loading task-analysis.md):**
+Does this task require reading code files to decide HOW to execute it?
+- **NO** → Execute directly, skip task-analysis.md
+- **YES** → Load and follow task-analysis.md
+
+**If not early exit, start here for any user request:**
 1. Apply `.agents/tasks/task-analysis.md`
 2. Follow its output to select the appropriate path:
 

package/README.md

@@ -4,7 +4,7 @@ Your AI (LLM), guided by built-in workflows. Describe what you want, and it foll
 
 [![MIT License](https://img.shields.io/badge/License-MIT-green.svg)](LICENSE)
 [![AGENTS.md](https://img.shields.io/badge/AGENTS.md-compliant-blue.svg)](https://agents.md)
-[![Version](https://img.shields.io/badge/version-0.5.1-blue.svg)](package.json)
+[![Version](https://img.shields.io/badge/version-0.6.2-blue.svg)](package.json)
 
 ![Demo: Building a Slack bot with Agentic Code](.github/assets/demo.gif)
 
@@ -110,11 +110,20 @@ cp -r path/to/agentic-code/.agents .
 
 `.agents/skills/` contains reusable skill files in the [Codex Skills format](https://github.com/openai/codex/blob/main/docs/skills.md). Each skill has a `SKILL.md` with instructions that AI agents can discover and apply.
 
-**Codex**: You can install skills globally so Codex picks them up across all projects:
+**Codex**: Install skills for Codex CLI:
 
 ```bash
+# User scope (all projects)
 npx agentic-code-install-skills --codex
 # Installs to ~/.codex/skills/agentic-code/
+
+# Project scope (current project only)
+npx agentic-code-install-skills --codex --project
+# Installs to ./.codex/skills/agentic-code/
+
+# Custom path
+npx agentic-code-install-skills --path ./custom/skills
+# Installs to ./custom/skills/agentic-code/
 ```
 
 ## Common Questions

package/bin/install-codex-skills.js

@@ -7,10 +7,28 @@ const os = require('os');
 // Parse command line arguments
 const args = process.argv.slice(2);
 
+// Scope types
+const SCOPES = {
+  user: 'user',
+  project: 'project'
+};
+
+// Get target directory based on target and scope
+function getTargetDir(targetKey, scope) {
+  if (targetKey === 'codex') {
+    if (scope === SCOPES.project) {
+      return path.join(process.cwd(), '.codex', 'skills');
+    }
+    // User scope: respect CODEX_HOME environment variable
+    const codexHome = process.env.CODEX_HOME || path.join(os.homedir(), '.codex');
+    return path.join(codexHome, 'skills');
+  }
+  return null;
+}
+
 // Target configurations
 const TARGETS = {
   codex: {
-    dir: path.join(os.homedir(), '.codex', 'skills'),
     name: 'agentic-code',
     description: 'OpenAI Codex CLI',
     postInstall: 'Restart Codex to load the skills.\nNote: Enable skills with --enable skills flag or in config.toml'
@@ -24,26 +42,70 @@ if (args.includes('--help') || args.includes('-h')) {
 Usage: agentic-code-install-skills [options]
 
 Options:
-  --codex    Install skills to ~/.codex/skills/.agentic-code (default)
-  --help     Show this help message
+  --codex         Install skills to Codex CLI (default)
+  --project       Install to project scope instead of user scope
+  --path <path>   Install to custom path
+  --help          Show this help message
+
+Scopes:
+  user (default)  Install to user directory (~/.codex/skills/ or $CODEX_HOME/skills/)
+  project         Install to current project (./.codex/skills/)
 
 Examples:
-  agentic-code-install-skills           # Install to Codex (default)
-  agentic-code-install-skills --codex   # Install to Codex (explicit)
+  agentic-code-install-skills                    # Install to ~/.codex/skills/ (user scope)
+  agentic-code-install-skills --codex            # Same as above (explicit)
+  agentic-code-install-skills --codex --project  # Install to ./.codex/skills/ (project scope)
+  agentic-code-install-skills --path ./my-skills # Install to ./my-skills/agentic-code
 `);
   process.exit(0);
 }
 
+// Parse --path option
+function getCustomPath() {
+  const pathIndex = args.indexOf('--path');
+  if (pathIndex === -1) return null;
+
+  const customPath = args[pathIndex + 1];
+  if (!customPath || customPath.startsWith('-')) {
+    console.error('Error: --path requires a path argument');
+    process.exit(1);
+  }
+  return path.resolve(customPath);
+}
+
+const customPath = getCustomPath();
+
 // Determine target (default: codex)
-let targetKey = 'codex';
+let targetKey = customPath ? 'custom' : 'codex';
 if (args.includes('--codex')) {
-  targetKey = 'codex';
+  targetKey = customPath ? 'custom' : 'codex';
 }
 
-const target = TARGETS[targetKey];
-if (!target) {
-  console.error(`Error: Unknown target '${targetKey}'`);
-  process.exit(1);
+// Determine scope (default: user)
+const scope = args.includes('--project') ? SCOPES.project : SCOPES.user;
+
+// Get skills directory
+let skillsDir;
+let target;
+
+if (customPath) {
+  skillsDir = customPath;
+  target = {
+    name: 'agentic-code',
+    description: 'custom path',
+    postInstall: null
+  };
+} else {
+  target = TARGETS[targetKey];
+  if (!target) {
+    console.error(`Error: Unknown target '${targetKey}'`);
+    process.exit(1);
+  }
+  skillsDir = getTargetDir(targetKey, scope);
+  if (!skillsDir) {
+    console.error(`Error: Cannot determine target directory for '${targetKey}'`);
+    process.exit(1);
+  }
 }
 
 // Get source directory (from installed package or local)
@@ -89,16 +151,17 @@ function main() {
     process.exit(1);
   }
 
-  const targetDir = path.join(target.dir, target.name);
+  const targetDir = path.join(skillsDir, target.name);
+  const scopeLabel = customPath ? 'custom' : (scope === SCOPES.project ? 'project' : 'user');
 
-  console.log(`Installing Agentic Code skills to ${target.description}...\n`);
+  console.log(`Installing Agentic Code skills to ${target.description} (${scopeLabel})...\n`);
   console.log(`Source: ${sourceDir}`);
   console.log(`Target: ${targetDir}\n`);
 
   // Create target parent directory if not exists
-  if (!fs.existsSync(target.dir)) {
-    fs.mkdirSync(target.dir, { recursive: true });
-    console.log(`Created: ${target.dir}`);
+  if (!fs.existsSync(skillsDir)) {
+    fs.mkdirSync(skillsDir, { recursive: true });
+    console.log(`Created: ${skillsDir}`);
   }
 
   // Remove existing target directory

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "agentic-code",
-  "version": "0.6.0",
+  "version": "0.6.2",
   "description": "Task-oriented context engineering framework for LLM coding agents - AGENTS.md standard compliant",
   "files": [
     "bin/",