undefined

undefined

undefined

```bash

**Note**: `agent-worker new` only creates standalone agents in the global workflow. Use YAML for workflow agents.

**Target syntax**:
- `alice` → standalone (`alice@global:main`)
- `alice@review` → agent in review workflow (`alice@review:main`)
- `alice@review:pr-123` → full specification
- `@review` → workflow reference (for broadcast/listing)
- `@review:pr-123` → specific workflow instance

**Context isolation**:

undefined

undefined

undefined

agent-worker new -m anthropic/claude-sonnet-4-5  # SDK (default)
agent-worker new -b claude                       # Claude CLI
agent-worker new -b cursor                       # Cursor Agent
agent-worker new -b mock                         # Testing (no API)

agent-worker new -b mock
agent-worker send a0 "Hello"

agent-worker new monitor --wakeup 30s --prompt "Check CI status"

undefined

```bash

---

undefined

```bash

undefined

undefined

tools: [bash, read, write]  # CLI backend tool names
max_tokens: 8000
max_steps: 20

# OR persistent - survives shutdown
bind: ./data/${{ workflow.tag }}/

undefined

setup:
  - shell: echo "pr-${{ env.PR_NUMBER }}"
    as: branch_name

kickoff: |
  Workflow: ${{ workflow.name }}
  Tag: ${{ workflow.tag }}
  Branch: ${{ branch_name }}

kickoff: |
  @alice Start the task.

kickoff: |
  @alice @bob @charlie All review this code.

agents:
  researcher:
    system_prompt: Research and write findings to the shared document.

  summarizer:
    system_prompt: Read the document and create a concise summary.

context:
  provider: file
  config:
    bind: ./results/  # Persistent across runs

undefined

```bash
PR_NUMBER=123 agent-worker run review.yaml --tag pr-123

undefined

```bash
TOPIC="AI agent frameworks" agent-worker run research.yaml

undefined

**Consensus Decision:**
```yaml

---

undefined

**Available tools** (in agent's system prompt):
- `channel_send` - Send message to channel
- `channel_read` - Read recent messages
- `inbox_read` - Read own @mentions

undefined

**Available tools** (in agent's system prompt):
- `document_read` - Read current document
- `document_write` - Overwrite document
- `document_append` - Append to document

{
  "name": "proposal_create",
  "arguments": {
    "title": "Merge PR #123",
    "type": "approval",
    "resolution": "majority"
  }
}

undefined

---

undefined

undefined

// my-tools.ts
export default [
  {
    name: 'search_docs',
    description: 'Search documentation',
    parameters: {
      type: 'object',
      properties: {
        query: { type: 'string', description: 'Search query' }
      },
      required: ['query']
    },
    needsApproval: false,  // Optional: require approval before execution
    execute: async (args) => {
      return { results: ['doc1', 'doc2'] }
    }
  }
]

undefined

undefined

agent-worker send a0 "Delete /tmp/test.txt"
agent-worker pending
agent-worker approve <id>
agent-worker deny <id> -r "Path not allowed"

undefined

Check available providers:
```bash
agent-worker providers

import { AgentSession } from 'agent-worker'

const session = new AgentSession({
  model: 'anthropic/claude-sonnet-4-5',
  system: 'You are a helpful assistant.',
  tools: [/* your tools */]
})

// Send message
const response = await session.send('Hello')
console.log(response.content)
console.log(response.toolCalls)
console.log(response.usage)

// Stream response
for await (const chunk of session.sendStream('Tell me a story')) {
  process.stdout.write(chunk)
}

// State management
const state = session.getState()
// Later: restore from state

import { AgentSession, SkillsProvider, createSkillsTool } from 'agent-worker'

const skillsProvider = new SkillsProvider()
await skillsProvider.scanDirectory('.agents/skills')

const session = new AgentSession({
  model: 'anthropic/claude-sonnet-4-5',
  system: 'You are a helpful assistant.',
  tools: [createSkillsTool(skillsProvider)]
})

undefined

---