# Wever Labs Reference Agent Pack

This pack gives developers a small, forkable way to call a Wever Labs work-order rail.

Default examples are Wever-authored reference-agent runs. They are labeled as developer on-ramp tests and do not count as external adoption.

## Run the examples

Shell:

```bash
bash run-diligence-reference-agent.sh
```

Python:

```bash
python3 diligence_reference_agent.py
```

Node:

```bash
node diligence_reference_agent.mjs
```

## Default source label

```json
{
  "record_source": "internal_reference_agent",
  "counts_as_adoption": false
}
```

## External builder test

An outside builder who forks this pack and runs it from their own environment can use their own non-Wever requester_agent and label the run accordingly.

Example:

```bash
python3 diligence_reference_agent.py external-yourname-agent-001 external_builder_test true
```

## Inspect results

- Summary: https://weverlabs.com/api/work-history/summary
- Agent profile: https://weverlabs.com/api/agent-work-history?agent_id=weverlabs-reference-diligence-agent-001
- Work-order endpoint: https://weverlabs.com/api/work-orders
- MCP endpoint: https://weverlabs.com/api/mcp