deep-researcher
Deep-researcher conducts thorough investigations of complex topics by performing initial web searches, extracting full-text content from multiple sources, cross-referencing claims across them, and generating formatted reports with citations. Use this skill when you need comprehensive, multi-source research synthesized into a detailed report rather than quick answers from a single source.
git clone --depth 1 https://github.com/clawdotnet/openclaw.net /tmp/deep-researcher && cp -r /tmp/deep-researcher/src/OpenClaw.Gateway/skills/deep-researcher ~/.claude/skills/deep-researcherSKILL.md
When asked to "research", "deep dive", or investigate a complex topic: 1) Initial Search Phase: - Use the `web_search` tool to find broad overviews of the topic. - Identify 3-5 high-quality, reputable sources. 2) Deep Extraction Phase: - Use the `web_fetch` tool to read the full text of the identified sources. - If documents are PDFs, use the `pdf_read` tool to extract their contents. - Take notes using the `memory` tool if the context is too large to hold at once. 3) Synthesis & Cross-Referencing: - Compare claims across different sources to verify accuracy. - Look for consensus, controversies, or primary data. 4) Report Generation: - Draft a comprehensive report with clear sections (e.g., Executive Summary, Deep Dive, Sources). - Use markdown formatting, tables, and bullet points to make data digestible. - Explicitly cite your sources with URLs. - Use the `write_file` tool to save the final report to the workspace if requested, or return it in the chat.
Organizes repository documentation and keeps new docs in the correct location.
Produce a daily news brief (AI/security/dev) with links, key takeaways, and action items.
Connects to databases, runs SQL queries, and analyzes datasets using code to provide actionable business insights.
Triage inbox into “needs reply”, “FYI”, “urgent”, and “safe to archive”, with a short plan for each.
Safely operate Home Assistant by resolving targets first and validating state before writes.
Safely interact with MQTT topics with allow/deny policies and minimal payload risk.
Operates as an autonomous software engineer, capable of writing code, running tests, and managing git repositories.