Manus Agent
Programmatic Manus API integration for deep research / autonomous task execution.
Status
- Verified live on 2026-06-04.
- Endpoint:
https://api.manus.ai/v1 - Auth:
x-manus-api-key: $MANUS_API_KEY(important: NOTAuthorization: Bearer; Bearer is OAuth only). - Model observed:
manus-1.6-agent. - End-to-end test passed: create task + poll completion.
Usage
bash
. /data/.openclaw/secrets/env.sh
cd /data/.openclaw/workspace/skills/manus-agent
scripts/manus.py --check
scripts/manus.py --ask "Research topic..." --timeout 600
scripts/manus.py --task <task_id>
When to use
- Autonomous multi-step research, report generation, task execution.
- Good fallback/parallel agent alongside Gemini Deep Research, Perplexity, Hermes.
- For high-stakes answers, cross-check Manus output with at least one other source/tool.
Notes
- API keys are created at Manus API Integration settings: https://manus.im/app?show_settings=integrations&app_name=api
- Each API key has broad Manus account access; rotate if exposed.