feat: add automated versioning and PyPI publishing

## Release Automation
- Integrated Release Please for PR-based versioning
- Automated PyPI publishing on release
- Conventional Commits for version bumps
- CHANGELOG generation

## Conductor Integration
- Added .conductor.json for automatic environment setup
- Simplified documentation with auto-setup
- No manual setup needed in new worktrees

## Security & Cleanup
- Removed private agent credentials from examples
- Deleted temporary documentation files
- Created generic .env.example without secrets
- Removed integration tests with private tokens

## Documentation
- RELEASING.md: Complete versioning and release guide
- CONDUCTOR.md: Simplified setup instructions
- Updated .env.example with generic examples only

## Release Please Configuration
- Version bumping per PR (not per commit)
- Automatic CHANGELOG generation
- PyPI publishing on release
- Conventional commit format required

Next: Set PYPI_API_TOKEN secret in GitHub for publishing

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>

Author: bokelley

.conductor.json

@@ -0,0 +1,18 @@
+{
+  "name": "adcp-client-python",
+  "description": "Python client for Ad Context Protocol (AdCP)",
+  "setup": {
+    "commands": [
+      {
+        "name": "setup-env",
+        "description": "Copy .env configuration from repository root",
+        "command": "python3 scripts/setup_conductor_env.py",
+        "runOnCreate": true
+      }
+    ]
+  },
+  "python": {
+    "version": "3.10+",
+    "install": "pip install -e .[dev]"
+  }
+}

.env.example

@@ -5,24 +5,10 @@
 CREATIVE_AGENT_URI=https://creative.adcontextprotocol.org
 CREATIVE_AGENT_PROTOCOL=mcp
 
-# Optable Signals Agent (MCP)
-OPTABLE_AGENT_URI=https://sandbox.optable.co/admin/adcp/signals/mcp
-OPTABLE_AGENT_PROTOCOL=mcp
-OPTABLE_AGENT_TOKEN=5ZWQoDY8sReq7CTNQdgPokHdEse8JB2LDjOfo530_9A=
-
-# Wonderstruck Sales Agent (MCP)
-WONDERSTRUCK_AGENT_ID=principal_8ac9e391
-WONDERSTRUCK_AGENT_NAME=Wonderstruck (MCP)
-WONDERSTRUCK_AGENT_URI=https://wonderstruck.sales-agent.scope3.com/mcp/
-WONDERSTRUCK_AGENT_PROTOCOL=mcp
-WONDERSTRUCK_AGENT_TOKEN=UhwoigyVKdd6GT8hS04cc51ckGfi8qXpZL6OvS2i2cU
-
-# Test Agent (A2A) - Currently returning 404 errors
-TEST_AGENT_ID=principal_3bd0d4a8
-TEST_AGENT_NAME=Test Agent
-TEST_AGENT_URI=https://test-agent.adcontextprotocol.org
-TEST_AGENT_PROTOCOL=a2a
-TEST_AGENT_TOKEN=L4UCklW_V_40eTdWuQYF6HD5GWeKkgV8U6xxK-jwNO8
+# Example Agent (MCP) - Replace with your agent details
+EXAMPLE_AGENT_URI=https://your-agent.example.com/mcp
+EXAMPLE_AGENT_PROTOCOL=mcp
+EXAMPLE_AGENT_TOKEN=your-token-here
 
 # Multi-Agent Configuration (JSON format)
 # This is the format expected by ADCPMultiAgentClient.from_env()
@@ -34,25 +20,11 @@ ADCP_AGENTS='[
     "protocol": "mcp"
   },
   {
-    "id": "optable_signals",
-    "name": "Optable Signals",
-    "agent_uri": "https://sandbox.optable.co/admin/adcp/signals/mcp",
+    "id": "your_agent",
+    "name": "Your Agent",
+    "agent_uri": "https://your-agent.example.com",
     "protocol": "mcp",
-    "auth_token": "5ZWQoDY8sReq7CTNQdgPokHdEse8JB2LDjOfo530_9A="
-  },
-  {
-    "id": "principal_8ac9e391",
-    "name": "Wonderstruck (MCP)",
-    "agent_uri": "https://wonderstruck.sales-agent.scope3.com/mcp/",
-    "protocol": "mcp",
-    "auth_token": "UhwoigyVKdd6GT8hS04cc51ckGfi8qXpZL6OvS2i2cU"
-  },
-  {
-    "id": "principal_3bd0d4a8",
-    "name": "Test Agent",
-    "agent_uri": "https://test-agent.adcontextprotocol.org",
-    "protocol": "a2a",
-    "auth_token": "L4UCklW_V_40eTdWuQYF6HD5GWeKkgV8U6xxK-jwNO8"
+    "auth_token": "your-token-here"
   }
 ]'
 

.github/workflows/release-please.yml

@@ -0,0 +1,48 @@
+name: Release Please
+
+on:
+  push:
+    branches:
+      - main
+
+permissions:
+  contents: write
+  pull-requests: write
+
+jobs:
+  release-please:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: googleapis/release-please-action@v4
+        id: release
+        with:
+          release-type: python
+          package-name: adcp
+
+      # Publish to PyPI when a release is created
+      - name: Checkout
+        if: ${{ steps.release.outputs.release_created }}
+        uses: actions/checkout@v4
+
+      - name: Setup Python
+        if: ${{ steps.release.outputs.release_created }}
+        uses: actions/setup-python@v5
+        with:
+          python-version: '3.10'
+
+      - name: Install build dependencies
+        if: ${{ steps.release.outputs.release_created }}
+        run: |
+          python -m pip install --upgrade pip
+          pip install build twine
+
+      - name: Build package
+        if: ${{ steps.release.outputs.release_created }}
+        run: python -m build
+
+      - name: Publish to PyPI
+        if: ${{ steps.release.outputs.release_created }}
+        env:
+          TWINE_USERNAME: __token__
+          TWINE_PASSWORD: ${{ secrets.PYPI_API_TOKEN }}
+        run: twine upload dist/*

.release-please-manifest.json

@@ -0,0 +1,3 @@
+{
+  ".": "0.1.0"
+}

CONDUCTOR.md

@@ -0,0 +1,108 @@
+# Conductor Setup
+
+## Automatic Setup
+
+When you create a new Conductor worktree, the environment is automatically configured via `.conductor.json`.
+
+The setup script copies `.env` from the repository root to your worktree, giving you instant access to configured agents.
+
+## First-Time Setup
+
+### 1. Configure Repository Root
+
+In the **repository root** (not in a worktree):
+
+```bash
+cd /path/to/adcp-client-python
+cp .env.example .env
+# Edit .env with your agent configurations
+```
+
+### 2. Start Working
+
+That's it! When you create or enter a Conductor worktree, the setup happens automatically.
+
+## Environment Variables
+
+### ADCP_AGENTS (Recommended)
+
+Use this JSON format for multi-agent configuration:
+
+```bash
+ADCP_AGENTS='[
+  {
+    "id": "creative_agent",
+    "name": "Creative Agent",
+    "agent_uri": "https://creative.adcontextprotocol.org",
+    "protocol": "mcp"
+  },
+  {
+    "id": "my_agent",
+    "name": "My Agent",
+    "agent_uri": "https://my-agent.example.com",
+    "protocol": "mcp",
+    "auth_token": "your-token-here"
+  }
+]'
+```
+
+Then in code:
+
+```python
+from adcp import ADCPMultiAgentClient
+
+client = ADCPMultiAgentClient.from_env()
+agent = client.agent("my_agent")
+```
+
+### Individual Variables
+
+For single-agent usage:
+
+```bash
+AGENT_URI=https://agent.example.com
+AGENT_PROTOCOL=mcp
+AGENT_TOKEN=your-token
+```
+
+## Manual Setup
+
+If automatic setup doesn't work, run manually:
+
+```bash
+python3 scripts/setup_conductor_env.py
+```
+
+## Security
+
+- ✅ `.env` is gitignored
+- ✅ Store tokens in `.env` at repository root
+- ✅ Use `.env.example` for documentation
+- ❌ Never commit actual tokens to git
+
+## Testing
+
+```bash
+# Creative agent (public, no auth)
+python3 tests/integration/test_creative_agent.py
+
+# Your own agents (configured in .env)
+python3 tests/integration/test_your_agent.py
+```
+
+## Troubleshooting
+
+**"No .env file found"**:
+```bash
+cd /path/to/adcp-client-python  # Repository root
+cp .env.example .env
+# Edit .env
+```
+
+**Setup didn't run automatically**:
+```bash
+# Run manually in your worktree
+python3 scripts/setup_conductor_env.py
+```
+
+See full documentation in README.md for complete usage examples.