\r \r

Deep Research for OpenClaw\r

\r

What this skill is\r

This is an integration skill for installing and wiring the deep-researcher OpenClaw sub-agent from the public repository:\r \r

• deep-research-openclaw-agent\r \r The repository contains:\r
• the workspace-researcher prompt pack;\r
• the local research helper scripts;\r
• the Main -> Deep Research orchestration contract;\r
• the report lint, validation, and finalization pipeline.\r \r This skill is intended for OpenClaw users who want a reproducible deep-research workflow without assembling the runtime and contracts from scratch.\r \r

What it can do\r

• structured deep research through plan -> scout -> harvest -> verify -> synthesize;\r
• hybrid discovery with web_search, Tavily, and web_fetch;\r
• explicit source registry, claim ledger, and coverage tracking;\r
• report linting, validation, and final M2M JSON finalization;\r
• honest SUCCESS | PARTIAL | FAILURE delivery with explicit gaps and conflicts.\r \r

Requirements\r

• OpenClaw 2026.3.x or later\r
• Python available on the host\r
• a configured deep-researcher agent in OpenClaw\r
• Tavily API access if you want the Tavily-backed path\r \r

Install\r

1. Clone the repository:\r
   • git clone https://github.com/MilleniumGenAI/deep-research-openclaw-agent.git\r
2. Copy openclaw/workspace-researcher/ into your OpenClaw base directory, or point your agent config at that path directly.\r
3. Align the main-agent handoff with:\r
   • openclaw/main-deep-research-skill.md\r
4. Register or update the deep-researcher agent in openclaw.json.\r
5. If you want Tavily-backed scouting, ensure TAVILY_API_KEY is available in env or .env.\r \r

Validate\r

Run these checks before using the agent in real work:\r \r

python -m py_compile openclaw/workspace-researcher/scripts/*.py\r
python openclaw/workspace-researcher/scripts/init_research_run.py --workspace openclaw/workspace-researcher --topic "Smoke test" --language en --task-date 2026-03-10\r
```\r
\r
Then run a first smoke task through OpenClaw once the agent is wired:\r
\r
```bash\r
openclaw agent --agent deep-researcher --json --message "Perform deep research using your local SOUL.md contract. GOAL: confirm the runtime can initialize a fresh run and return PARTIAL if no external research is performed. SCOPE: in scope is only local init and artifact creation; out of scope is web research. SUCCESS CRITERIA: create fresh tmp artifacts and explain blocked evidence collection honestly. TASK DATE: 2026-03-10. DELIVERABLES: finalized M2M JSON. LANGUAGE: en. CONSTRAINTS: do not fabricate sources; return PARTIAL if evidence is insufficient."\r
```\r
\r
## Core references\r
- Root README: [README.md](https://github.com/MilleniumGenAI/deep-research-openclaw-agent/blob/main/README.md)\r
- Sub-agent contract: [openclaw/workspace-researcher/SOUL.md](https://github.com/MilleniumGenAI/deep-research-openclaw-agent/blob/main/openclaw/workspace-researcher/SOUL.md)\r
- Main handoff contract: [openclaw/main-deep-research-skill.md](https://github.com/MilleniumGenAI/deep-research-openclaw-agent/blob/main/openclaw/main-deep-research-skill.md)\r
- Runtime scripts: [openclaw/workspace-researcher/scripts/](https://github.com/MilleniumGenAI/deep-research-openclaw-agent/tree/main/openclaw/workspace-researcher/scripts)\r
- Agent config template: [openclaw/agent-config.template.json](https://github.com/MilleniumGenAI/deep-research-openclaw-agent/blob/main/openclaw/agent-config.template.json)\r
- Known limits: [docs/known-limits.md](https://github.com/MilleniumGenAI/deep-research-openclaw-agent/blob/main/docs/known-limits.md)\r
\r
## Notes\r
- This is an OpenClaw-only v1 package.\r
- ClawHub publishes skills under platform-wide MIT-0 terms.\r
- The runtime source of truth is `openclaw/workspace-researcher/SOUL.md`.\r
- Findings should be built only from traceable external sources, not from local artifacts.\r