grounded-research-lit
Run focused literature and web research from a grounded note. Use when a grounded note already exists and you want targeted research results, opened-link evidence, deeper per-paper analysis materials, optional downloaded literature, and a two-stage literature output (`lit_initial.md` then refined `lit.md`).
Profile →npx skills add gaotiexinqu/OneResearchClaw --skill grounded-research-lit --agent cursor
Same command for any agent — swap --agent for claude-code, codex, copilot.
Weekly change comes from our own snapshots, not the repository page — it measures attention, not adoption.
What it does
Guides a research agent to perform targeted literature and web research based on an existing grounded note. It adapts to the grounded note schema, creates focused query groups, and executes search via either an external API backend or Cursor-native search/browse. It preserves opened-source materials, builds structured per-paper notes, and produces an initial literature report (lit_initial.md) followed by a refined final report (lit.md). It can optionally download opened literature and refines notes after PDFs are downloaded.
How it works
- Infer ground_id from the grounded note path and read the grounded note. 2. Detect the grounding type from the note structure. 3. Load constants from config/research_pipeline.env. 4. If queries_confirmed_path is provided, read queries from that file and copy them as queries.json. Otherwise, extract main topic, open questions, constraints/risks, likely method directions, and keywords from the grounded note and generate three query groups: problem/background, method/solution, and domain/constraint; write these as plain string arrays in queries.json. 5. Run research using either the external backend (web_search_reader.py) or Cursor-native search/browse. 6. Preserve opened readable sources under opened_sources/. 7. Generate prepared per-paper notes using prepare_opened_paper_notes.py and write opened_paper_notes.jsonl. 8. If backend is Cursor-native, verify opened-paper count against MIN_OPENED_PAPERS and recent papers rule before writing lit_initial.md. 9. Perform Paper Opened Confirmation inline: for each opened paper, check for a corresponding opened source with ≥ 500 readable chars and substantive analysis; if not, re-search and re-open, regenerating notes; remove papers that cannot reach sufficient content. 10. Write data/lit_inputs/<ground_id>/lit_initial.md from opened-page evidence and paper notes. 11. If DOWNLOAD_OPENED_LITERATURE=true, run downloads (externally or via download_opened_literature.py) as an auxiliary stage between lit_initial.md and lit.md. 12. Wait for the download manifest to reach terminal state. 13. Run refine_notes_from_downloaded_pdfs.py to enrich notes from PDFs. 14. Write the final lit.md from refined notes.
When to use it
Use when a grounded note exists and you want targeted literature results with evidence and two-stage outputs, optionally with downloaded literature.
What it can touch
- Cursor-native and external backends (controlled via config/research_pipeline.env). Tools involved include: web_search_reader.py, prepare_opened_paper_notes.py, download_opened_literature.py, refine_notes_from_downloaded_pdfs.py. Output artifacts are written under data/lit_inputs/<ground_id>/ and data/lit_results/<ground_id>/.
Caveats
- The workflow includes an inline Paper Opened Confirmation step to ensure opened sources have sufficient content and substantive notes; if not resolvable, papers are dropped from lit_initial.md. - All file paths and script names are used exactly as described in the skill. - No guarantees about external backend results or download success beyond described rules.
# Grounded Research Literature Use a structured grounded note to perform **targeted literature / web research**. This skill is **not** a generic survey skill. It is a **grounding-conditioned research stage**. Its responsibilities are: 1. read a grounded note 2. adapt to the grounded note's schema 3. generate focused search queries 4. save them to a JSON file 5. run research using either: - externa
What does the grounded-research-lit skill do?
Run focused literature and web research from a grounded note. Use when a grounded note already exists and you want targeted research results, opened-link evidence, deeper per-paper analysis materials, optional downloaded literature, and a two-stage literature output (`lit_initial.md` then refined `lit.md`).
How do I install it?
Run `npx skills add gaotiexinqu/OneResearchClaw --skill grounded-research-lit --agent claude-code` — it drops the skill into your project so the agent can pick it up. Swap the --agent value for codex, cursor or copilot if you use one of those.
Where does this skill come from?
From gaotiexinqu/OneResearchClaw, a repository with 438 stars. We read it straight from the repository tree rather than a submitted listing, so what you see here is what is actually published.
Is a popular skill a good skill?
Not necessarily. Stars measure attention, not adoption — a repository can trend for a week and be abandoned. That is why we show the weekly change from our own snapshots next to the total, instead of a single flattering number.