prjct-cli 0.33.5 → 0.34.0

package/CHANGELOG.md

@@ -1,5 +1,70 @@
 # Changelog
 
+## [0.34.0] - 2026-01-15
+
+### Feature: Agentic Orchestrator + MCP Auto-Install
+
+Major release with intelligent task routing and simplified integration setup.
+
+#### Orchestrator Agent (NEW)
+
+Auto-routing system for tasks to the right agents and skills:
+
+- **Domain Detection**: Analyzes task keywords to detect domains (frontend, backend, database, etc.)
+- **Agent Loading**: Automatically loads relevant agents from `{globalPath}/agents/`
+- **Skill Invocation**: Invokes skills from `~/.claude/skills/` based on task context
+- **Multi-Domain Coordination**: Handles tasks spanning multiple domains with proper execution order
+
+**Usage**: Orchestrator runs automatically for task-related commands (`p. task`, `p. done`, `p. ship`, etc.)
+
+#### MCP Server Auto-Install
+
+All integrations now auto-install their MCP servers:
+
+- **Linear**: Auto-installs `mcp-remote` for Linear API
+- **JIRA**: Auto-installs Atlassian MCP server
+- **GitHub**: Auto-installs GitHub MCP server with token validation
+- **Monday.com**: Auto-installs Monday MCP server
+
+No more manual `~/.claude/settings.json` editing - just run `p. linear setup` and it handles everything.
+
+#### Individual Command Skills
+
+All 37 prjct commands now available as individual Claude Code skills:
+
+- `/p.task` - Start a task
+- `/p.sync` - Sync project
+- `/p.linear` - Linear integration
+- Plus 34 more commands
+
+#### agentskills.io Integration
+
+Skills are now discovered from the official agentskills.io ecosystem:
+
+- Skills from `anthropics/skills` GitHub repo
+- Proper SKILL.md format with YAML frontmatter
+- Agent-to-skill mapping in `templates/config/skill-mappings.json`
+
+**Files Changed:**
+- `templates/agentic/orchestrator.md` (NEW)
+- `templates/commands/p.md` - Orchestrator integration
+- `templates/commands/task.md` - Orchestrator context usage
+- `templates/commands/linear.md` - MCP auto-install
+- `templates/commands/jira.md` - MCP auto-install
+- `templates/commands/github.md` - MCP auto-install
+- `templates/commands/monday.md` - MCP auto-install
+- `templates/config/skill-mappings.json` - agentskills.io sources
+- `scripts/postinstall.js` - Individual command installation
+
+**Removed (deprecated):**
+- `templates/commands/feature.md`
+- `templates/commands/now.md`
+- `templates/commands/dashboard.md`
+- `templates/commands/suggest.md`
+- `templates/commands/ask.md`
+
+---
+
 ## [0.33.5] - 2026-01-13
 
 ### Fix: Type Safety Improvements (PRJ-54)

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "prjct-cli",
-  "version": "0.33.5",
+  "version": "0.34.0",
   "description": "Built for Claude - Ship fast, track progress, stay focused. Developer momentum tool for indie hackers.",
   "main": "core/index.ts",
   "bin": {

package/scripts/postinstall.js

@@ -23,7 +23,7 @@ const VERSION = pkg.version
 
 console.log('\n   prjct-cli postinstall\n')
 
-// 1. Copy p.md router (CRITICAL - this is the only essential file)
+// 1. Copy p.md router (CRITICAL - main entry point)
 try {
   fs.mkdirSync(COMMANDS_DIR, { recursive: true })
 
@@ -41,7 +41,31 @@ try {
   console.log('   Run: npx prjct-cli setup')
 }
 
-// 2. Statusline (best-effort, not critical)
+// 2. Install individual commands as separate skills (p.task, p.sync, etc.)
+try {
+  const commandsDir = path.join(ROOT, 'templates', 'commands')
+  const commands = fs.readdirSync(commandsDir).filter(f => f.endsWith('.md') && f !== 'p.md')
+
+  let installed = 0
+  for (const cmd of commands) {
+    const src = path.join(commandsDir, cmd)
+    const cmdName = cmd.replace('.md', '')
+    const dest = path.join(COMMANDS_DIR, `p.${cmdName}.md`)
+
+    try {
+      fs.copyFileSync(src, dest)
+      installed++
+    } catch {
+      // Skip files that fail to copy
+    }
+  }
+
+  console.log(`   \u2713 ${installed} individual commands installed (/p.task, /p.sync, etc.)`)
+} catch (error) {
+  console.log('   ! Could not install individual commands:', error.message)
+}
+
+// 3. Statusline (best-effort, not critical)
 try {
   const STATUSLINE_SRC = path.join(ROOT, 'assets', 'statusline')
   const STATUSLINE_DEST = path.join(HOME, '.prjct-cli', 'statusline')

package/templates/agentic/orchestrator.md

@@ -0,0 +1,303 @@
+# Orchestrator - Agent & Skill Coordinator
+
+**Purpose**: Automatically route tasks to the right agents and skills.
+
+## How It Works
+
+```
+User Task → Orchestrator → [Analyze] → [Load Agents] → [Invoke Skills] → [Execute]
+```
+
+The orchestrator is **implicit** - it runs automatically for every p. command.
+
+---
+
+## Step 1: Task Analysis
+
+Analyze the current task/command to determine domains involved.
+
+### Domain Detection Keywords
+
+| Domain | Keywords | Agent | Skills |
+|--------|----------|-------|--------|
+| **Frontend** | react, vue, component, UI, CSS, styling, layout | `frontend.md` | ui-design, react-patterns |
+| **UX/UI** | design, user experience, accessibility, UX, interaction | `uxui.md` | ux-research, accessibility |
+| **Backend** | API, server, endpoint, database query, auth | `backend.md` | api-design, backend-patterns |
+| **Database** | schema, migration, query, SQL, ORM, prisma | `database.md` | sql-patterns, database-design |
+| **Testing** | test, spec, coverage, TDD, unit, integration | `testing.md` | test-automation |
+| **DevOps** | deploy, CI/CD, docker, kubernetes, pipeline | `devops.md` | ci-cd, infrastructure |
+| **Planning** | plan, architecture, design doc, PRD, spec | `prjct-planner.md` | architecture |
+| **Shipping** | ship, release, PR, merge, version | `prjct-shipper.md` | code-review |
+
+### Analysis Process
+
+```
+1. EXTRACT keywords from task description
+2. MATCH keywords to domains (can be multiple)
+3. SET: {detectedDomains} = matched domains
+4. SET: {primaryDomain} = highest match count
+5. SET: {secondaryDomains} = other matches
+```
+
+---
+
+## Step 2: Load Project Context
+
+```
+READ: .prjct/prjct.config.json → {projectId}
+SET: {globalPath} = ~/.prjct-cli/projects/{projectId}
+
+READ: {globalPath}/config/skills.json → {skillsConfig}
+READ: {globalPath}/analysis/repo-analysis.json → {repoAnalysis}
+```
+
+---
+
+## Step 3: Load Relevant Agents
+
+For each detected domain, load the corresponding agent.
+
+```
+SET: {loadedAgents} = []
+
+FOR EACH domain IN {detectedDomains}:
+  SET: {agentPath} = {globalPath}/agents/{domain}.md
+
+  IF file exists:
+    READ: {agentPath}
+    EXTRACT: frontmatter (description, skills, patterns)
+    ADD to {loadedAgents}
+
+    OUTPUT: "🤖 Loaded: {domain} agent"
+```
+
+### Agent Context Injection
+
+Each loaded agent provides:
+- **Patterns**: Code patterns specific to this project
+- **Conventions**: Naming, structure, style rules
+- **Skills**: Which skills to invoke for this domain
+- **Anti-patterns**: What to avoid
+
+---
+
+## Step 4: Invoke Skills
+
+For each loaded agent, check if skills should be invoked.
+
+```
+FOR EACH agent IN {loadedAgents}:
+  GET: {agentSkills} = agent.frontmatter.skills
+
+  FOR EACH skillName IN {agentSkills}:
+    SET: {skillPath} = ~/.claude/skills/{skillName}/SKILL.md
+
+    IF file exists:
+      READ: {skillPath}
+      EXTRACT: skill instructions
+      ADD to {activeSkills}
+
+      OUTPUT: "⚡ Skill active: {skillName}"
+```
+
+### Skill Selection Criteria
+
+Only invoke skills that are:
+1. **Relevant** to the current task
+2. **Installed** in ~/.claude/skills/
+3. **Linked** to a loaded agent
+
+---
+
+## Step 5: Build Execution Context
+
+Combine all context for task execution.
+
+```json
+{
+  "task": "{original task description}",
+  "command": "{p. command being executed}",
+  "project": {
+    "id": "{projectId}",
+    "ecosystem": "{repoAnalysis.ecosystem}",
+    "conventions": "{repoAnalysis.conventions}"
+  },
+  "agents": [
+    {
+      "name": "{agent.name}",
+      "patterns": "{agent.patterns}",
+      "rules": "{agent.rules}"
+    }
+  ],
+  "skills": [
+    {
+      "name": "{skill.name}",
+      "instructions": "{skill.instructions}"
+    }
+  ],
+  "execution": {
+    "primaryDomain": "{primaryDomain}",
+    "secondaryDomains": ["{secondaryDomains}"],
+    "commands": "{repoAnalysis.commands}"
+  }
+}
+```
+
+---
+
+## Step 6: Execute with Context
+
+Pass the execution context to the command template.
+
+The command template receives:
+- `{orchestrator.agents}` - Loaded agent contexts
+- `{orchestrator.skills}` - Active skill instructions
+- `{orchestrator.project}` - Project conventions
+- `{orchestrator.execution}` - Execution metadata
+
+---
+
+## Multi-Domain Coordination
+
+When task spans multiple domains:
+
+### Example: "Add user authentication with login form"
+
+```
+Detected domains:
+├── Frontend (login form, UI)
+├── Backend (auth API, sessions)
+└── Database (users table, schema)
+
+Loaded agents:
+├── frontend.md → React patterns, form handling
+├── backend.md → API design, auth patterns
+└── database.md → Schema design, migrations
+
+Active skills:
+├── ui-design → Form components
+├── api-design → Auth endpoints
+└── sql-patterns → User schema
+
+Execution order:
+1. Database: Create users schema
+2. Backend: Implement auth API
+3. Frontend: Build login form
+```
+
+---
+
+## Orchestrator Output
+
+At the start of each command, output:
+
+```
+🎯 Task: {task description}
+
+📦 Context Loaded:
+├── Agents: {loadedAgents.join(', ')}
+├── Skills: {activeSkills.join(', ')}
+└── Primary: {primaryDomain}
+
+{Continue with command execution...}
+```
+
+---
+
+## Integration with Commands
+
+Every p. command should:
+
+1. **Before execution**: Run orchestrator Steps 1-5
+2. **During execution**: Use orchestrator context
+3. **After execution**: Log which agents/skills were used
+
+### Command Template Integration
+
+```markdown
+# p. {command}
+
+## Step 0: Orchestrator
+
+INCLUDE: templates/agentic/orchestrator.md
+
+Execute orchestrator Steps 1-5 to build context.
+
+## Step 1: {Command-specific logic}
+
+Use {orchestrator.agents} and {orchestrator.skills} for:
+- Code patterns
+- Conventions
+- Domain expertise
+
+{Rest of command...}
+```
+
+---
+
+## Configuration
+
+### Disable Orchestrator
+
+For simple commands that don't need orchestration:
+
+```yaml
+---
+orchestrator: false
+---
+```
+
+### Force Specific Agents
+
+```yaml
+---
+orchestrator:
+  agents: [frontend, testing]
+  skills: [ui-design]
+---
+```
+
+---
+
+## Skill Invocation Patterns
+
+### When to Invoke Skills
+
+| Trigger | Skills to Invoke |
+|---------|------------------|
+| Creating UI component | ui-design, accessibility |
+| Writing API endpoint | api-design, backend-patterns |
+| Database changes | sql-patterns, database-design |
+| Writing tests | test-automation |
+| Deploying | ci-cd, infrastructure |
+| Code review | code-review |
+| Creating documents | pdf, docx, pptx (if installed) |
+
+### Skill Execution
+
+Skills provide:
+1. **Instructions** - How to approach the task
+2. **Examples** - Code/pattern examples
+3. **Checklists** - Quality gates
+4. **References** - Additional documentation
+
+---
+
+## Error Handling
+
+| Situation | Action |
+|-----------|--------|
+| No agents found | Use default patterns from repo-analysis |
+| No skills installed | Continue without skills, suggest `p. sync` |
+| Agent file missing | Skip that agent, continue with others |
+| Skill file missing | Skip that skill, log warning |
+
+---
+
+## Logging
+
+Log orchestrator decisions to memory:
+
+```json
+{"ts":"{timestamp}","action":"orchestrator","task":"{task}","agents":["{agents}"],"skills":["{skills}"],"primaryDomain":"{domain}"}
+```

package/templates/commands/bug.md

@@ -10,6 +10,8 @@ backend-sync: 'sync/pending.json'
 
 # /p:bug - Report Bug with Auto-Start
 
+**ARGUMENTS**: $ARGUMENTS
+
 ## Usage
 
 ```

package/templates/commands/github.md

@@ -5,6 +5,8 @@ description: 'GitHub Issues integration via MCP'
 
 # p. github - GitHub Issues Integration
 
+**ARGUMENTS**: $ARGUMENTS
+
 Manage GitHub Issues directly from prjct using MCP.
 
 ## Context Variables
@@ -50,17 +52,65 @@ IF file not found:
 
 ---
 
-## Step 2: Check MCP Tools
+## Step 2: Check GITHUB_TOKEN
+
+```
+CHECK: Is GITHUB_TOKEN env var set?
+
+IF not set:
+  OUTPUT: "GitHub requires a Personal Access Token."
+  OUTPUT: ""
+  OUTPUT: "1. Create token at: https://github.com/settings/tokens"
+  OUTPUT: "2. Select 'repo' scope"
+  OUTPUT: "3. Set env var: export GITHUB_TOKEN=ghp_..."
+  OUTPUT: "4. Run `p. github setup` again"
+  STOP
+```
+
+---
+
+## Step 3: Install MCP Server (if needed)
+
+```
+READ: ~/.claude/settings.json (create {} if not exists)
+CHECK: Does mcpServers.github exist?
+
+IF not exists:
+  READ: templates/mcp-config.json
+  EXTRACT: mcpServers.github
+
+  MERGE into ~/.claude/settings.json:
+  {
+    "mcpServers": {
+      "github": {
+        "command": "npx",
+        "args": ["-y", "@modelcontextprotocol/server-github"],
+        "env": {
+          "GITHUB_PERSONAL_ACCESS_TOKEN": "${GITHUB_TOKEN}"
+        }
+      }
+    }
+  }
+
+  WRITE: ~/.claude/settings.json
+
+  OUTPUT: "✅ Installed GitHub MCP server"
+  OUTPUT: ""
+  OUTPUT: "⚠️ Restart Claude Code to activate the MCP server."
+  OUTPUT: "Then run `p. github setup` again to complete configuration."
+  STOP
+```
+
+---
+
+## Step 4: Check MCP Tools Available
 
 ```
 CHECK: Are mcp__github__* tools available?
 
 IF not available:
-  OUTPUT: "GitHub MCP not configured."
-  OUTPUT: "1. Set GITHUB_TOKEN env var (needs repo scope)"
-  OUTPUT: "2. Add to ~/.claude/settings.json:"
-  OUTPUT: '{"mcpServers":{"github":{"command":"npx","args":["-y","@modelcontextprotocol/server-github"],"env":{"GITHUB_PERSONAL_ACCESS_TOKEN":"${GITHUB_TOKEN}"}}}}'
-  OUTPUT: "3. Restart Claude Code."
+  OUTPUT: "GitHub MCP is installed but not yet active."
+  OUTPUT: "Restart Claude Code, then run `p. github setup` again."
   STOP
 ```
 
@@ -70,7 +120,13 @@ IF not available:
 
 ### Flow
 
-1. **Verify MCP tools available**
+1. **Check token + Install MCP + Verify tools**
+   ```
+   Execute Step 2 (check GITHUB_TOKEN)
+   Execute Step 3 (auto-install MCP if needed)
+   Execute Step 4 (verify tools active)
+   IF not available: Prompt restart and STOP
+   ```
 
 2. **Detect repo from git remote**
    ```bash

package/templates/commands/jira.md

@@ -5,6 +5,8 @@ description: 'JIRA issue tracker integration'
 
 # p. jira - JIRA Integration
 
+**ARGUMENTS**: $ARGUMENTS
+
 Manage JIRA issues directly from prjct.
 
 ## Context Variables
@@ -68,7 +70,38 @@ IF file not found:
 
 ---
 
-## Step 2: Detect Auth Mode
+## Step 2: Install MCP Server (if needed)
+
+```
+READ: ~/.claude/settings.json (create {} if not exists)
+CHECK: Does mcpServers.Atlassian exist?
+
+IF not exists:
+  READ: templates/mcp-config.json
+  EXTRACT: mcpServers.Atlassian
+
+  MERGE into ~/.claude/settings.json:
+  {
+    "mcpServers": {
+      "Atlassian": {
+        "command": "npx",
+        "args": ["-y", "mcp-remote@latest", "https://mcp.atlassian.com/v1/sse"]
+      }
+    }
+  }
+
+  WRITE: ~/.claude/settings.json
+
+  OUTPUT: "✅ Installed Atlassian MCP server"
+  OUTPUT: ""
+  OUTPUT: "⚠️ Restart Claude Code to activate the MCP server."
+  OUTPUT: "Then run `p. jira setup` again to complete configuration."
+  STOP
+```
+
+---
+
+## Step 3: Detect Auth Mode
 
 ```
 # Check API Token first
@@ -79,6 +112,12 @@ IF JIRA_API_TOKEN is set:
 ELSE IF mcp__atlassian__jira_* tools available:
   SET: authMode = "mcp"
 
+# MCP installed but not active
+ELSE IF mcpServers.Atlassian exists in settings.json:
+  OUTPUT: "Atlassian MCP is installed but not yet active."
+  OUTPUT: "Restart Claude Code, then run `p. jira setup` again."
+  STOP
+
 # Neither available
 ELSE:
   OUTPUT: "JIRA not configured. Run `p. jira setup` first."
@@ -91,10 +130,11 @@ ELSE:
 
 ### Flow
 
-1. **Detect available auth methods**
+1. **Install MCP + Detect auth mode**
    ```
-   CHECK: JIRA_API_TOKEN env var
-   CHECK: Atlassian MCP tools available
+   Execute Step 2 (auto-install MCP if needed)
+   Execute Step 3 (detect auth mode)
+   IF MCP not active: Prompt restart and STOP
    ```
 
 2. **If MCP available (preferred)**

package/templates/commands/linear.md

@@ -5,6 +5,8 @@ description: 'Linear issue tracker integration via MCP'
 
 # p. linear - Linear Integration
 
+**ARGUMENTS**: $ARGUMENTS
+
 Manage Linear issues directly from prjct using MCP (no SDK needed).
 
 ## Context Variables
@@ -50,16 +52,45 @@ IF file not found:
 
 ---
 
-## Step 2: Check MCP Tools
+## Step 2: Install MCP Server (if needed)
+
+```
+READ: ~/.claude/settings.json (create {} if not exists)
+CHECK: Does mcpServers.linear exist?
+
+IF not exists:
+  READ: templates/mcp-config.json
+  EXTRACT: mcpServers.linear
+
+  MERGE into ~/.claude/settings.json:
+  {
+    "mcpServers": {
+      "linear": {
+        "command": "npx",
+        "args": ["-y", "mcp-remote", "https://mcp.linear.app/mcp"]
+      }
+    }
+  }
+
+  WRITE: ~/.claude/settings.json
+
+  OUTPUT: "✅ Installed Linear MCP server"
+  OUTPUT: ""
+  OUTPUT: "⚠️ Restart Claude Code to activate the MCP server."
+  OUTPUT: "Then run `p. linear setup` again to complete configuration."
+  STOP
+```
+
+---
+
+## Step 3: Check MCP Tools Available
 
 ```
 CHECK: Are mcp__linear__* tools available?
 
 IF not available:
-  OUTPUT: "Linear MCP not configured."
-  OUTPUT: "Add to ~/.claude/settings.json or .mcp.json:"
-  OUTPUT: '{"mcpServers":{"linear":{"command":"npx","args":["-y","mcp-remote","https://mcp.linear.app/mcp"]}}}'
-  OUTPUT: "Then restart Claude Code."
+  OUTPUT: "Linear MCP is installed but not yet active."
+  OUTPUT: "Restart Claude Code, then run `p. linear setup` again."
   STOP
 ```
 
@@ -69,10 +100,11 @@ IF not available:
 
 ### Flow
 
-1. **Verify MCP tools available**
+1. **Install MCP + Verify tools available**
    ```
-   CHECK: mcp__linear__* tools exist
-   IF not: Show MCP config instructions, STOP
+   Execute Step 2 (auto-install if needed)
+   Execute Step 3 (verify tools active)
+   IF not available: Prompt restart and STOP
    ```
 
 2. **Test connection (triggers OAuth if needed)**