@sidclaw/sdk

Governance for AI agents. Identity, policy, approval, and audit.

Documentation · Quick Start · Examples

Quick Start

1. Install

npm install @sidclaw/sdk

2. Initialize

import { AgentIdentityClient } from '@sidclaw/sdk';

const client = new AgentIdentityClient({
  apiKey: process.env.AGENT_IDENTITY_API_KEY,
  apiUrl: 'https://api.sidclaw.com',
  agentId: 'your-agent-id',
});

3. Govern your agent's actions

import { withGovernance } from '@sidclaw/sdk';

const sendEmail = withGovernance(client, {
  operation: 'send_email',
  target_integration: 'email_service',
  resource_scope: 'customer_emails',
  data_classification: 'confidential',
}, async (to, subject, body) => {
  // your actual send logic
});

await sendEmail('[email protected]', 'Follow-up', '...');
// If policy says "approval_required", this waits for a human to approve
// If policy says "deny", this throws ActionDeniedError
// If policy says "allow", this executes immediately

Integrations

MCP Governance Server

Wrap any MCP server with governance — intercepts tool calls automatically.

import { AgentIdentityClient, GovernanceMCPServer } from '@sidclaw/sdk';

const server = new GovernanceMCPServer({
  client,
  upstream: { transport: 'stdio', command: 'npx', args: ['your-mcp-server'] },
  toolMappings: [
    { toolName: 'query', data_classification: 'confidential' },
    { toolName: 'list_tables', skip_governance: true },
  ],
});
await server.start();

LangChain.js

import { governTools } from '@sidclaw/sdk/langchain';

const governedTools = governTools(myTools, { client });

Vercel AI SDK

import { governVercelTool } from '@sidclaw/sdk/vercel-ai';

const governedTool = governVercelTool('myTool', myTool, { client });

OpenAI Agents SDK

import { governOpenAITool } from '@sidclaw/sdk/openai-agents';

const governedTool = governOpenAITool(myTool, { client });

Webhook Verification

import { verifyWebhookSignature } from '@sidclaw/sdk/webhooks';

const isValid = verifyWebhookSignature(rawBody, signatureHeader, webhookSecret);

OpenClaw Integration

Add governance to any OpenClaw MCP server:

1. Install the skill: openclaw skills install sidclaw-governance
2. Replace your MCP server config in openclaw.json:

{
  "mcpServers": {
    "my-server": {
      "command": "npx",
      "args": ["-y", "@sidclaw/sdk", "mcp-proxy"],
      "env": {
        "SIDCLAW_API_KEY": "ai_...",
        "SIDCLAW_AGENT_ID": "agent-...",
        "SIDCLAW_UPSTREAM_CMD": "npx",
        "SIDCLAW_UPSTREAM_ARGS": "-y,@modelcontextprotocol/server-postgres,postgresql://..."
      }
    }
  }
}

Every tool call is now evaluated against your SidClaw policies.

Error Handling

import { ActionDeniedError, ApprovalTimeoutError } from '@sidclaw/sdk';

try {
  await governedAction();
} catch (error) {
  if (error instanceof ActionDeniedError) {
    console.log('Denied:', error.reason, 'Trace:', error.traceId);
  }
}

License

Apache 2.0