菜单
Identify research gaps and open questions from the literature in a workspace. Combines auditable search, citation analysis, and cross-paper comparison. Use when the user wants to find unexplored areas, formulate research questions, or assess where the field is heading.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
| name | research-gap |
| description | Identify research gaps and open questions from the literature in a workspace. Combines auditable search, citation analysis, and cross-paper comparison. Use when the user wants to find unexplored areas, formulate research questions, or assess where the field is heading. |
Systematically discover research gaps and open questions from the literature in a workspace.
The user must specify a workspace (--ws NAME) containing a sufficient number of papers (10+ recommended).
Report language is specified by the user (English or Chinese).
autor ws show <name> # paper list
autor ws search <name> "<broad theme>" # auditable node-level field scan
Perform an L2 scan (title + abstract) on workspace papers to build a field map.
autor ws search <name> "<theme keywords>"
autor research "<theme or gap question>" --run-dir workspace/<name>/runs/<case>
Are workspace papers concentrated in a few topics? Which related topics lack coverage?
Tally workspace papers by year to identify:
Use Python to read data/papers/*/meta.json directly for statistical analysis.
Scan L3/L4 of workspace papers to construct a methodology matrix:
autor show <dir_name> --level 3 # L3 gives paper-level takeaway, key findings, quantitative signals, and limitations when available
autor shared-refs "<id1>" "<id2>" # shared references
autor refs "<id>" # reference list
autor citing "<id>" # papers that cite this one
Load L3 of highly cited workspace papers and extract future directions, limitations, and unresolved questions. If L3 mode is inferred_synthesis, verify author-stated future work in L4 before treating it as the authors' own proposal:
autor show <dir_name> --level 3
Have those future directions been addressed? Cross-search to verify:
autor search "<future work keywords>"
Generate a structured research-gap report and save it to workspace/<name>/research-gaps.md.
Classify each identified gap by type:
| Gap Type | Meaning | Example |
|---|---|---|
| Knowledge gap | A phenomenon/question no one has studied yet | "The X effect at high Re has not been measured" |
| Methodological gap | Existing conclusions but method has flaws or limitations | "All current studies use RANS; DNS validation is lacking" |
| Contradiction gap | Different studies reach contradictory conclusions | "Group A reports a positive effect; Group B reports a negative one" |
| Transfer gap | A method/finding has not been extended to a related domain | "This method works in 2D; 3D has not been attempted" |
| Scale gap | Only small-scale or restricted results exist | "Only low-Re data; not validated at engineering Re" |
Report structure:
Quantitative support: when useful, write Python code to batch-extract data from meta.json, produce statistical charts (year distribution, method frequency, parameter-space coverage), and use visualizations to substantiate gap findings.
After the report is generated, proactively discuss with the user:
User says: "Find research gaps in the drag-review workspace." → Comprehensively scan the literature, multi-dimensional analysis, generate research-gap report
User says: "What future work directions did these papers mention? Has anyone followed up?" → Extract future-work, limitation, and unresolved-question signals from each paper's L3 card, then cross-search and verify in L4 when needed
Adversarial but evidence-bound Reviewer
Plan a literature review before drafting. Build a lean evidence-bound, taste-gated package using references.bib, reference-map.json, review-plan.md, evidence-ledger.md, and table-figure-plan.md.
Draft a fact-grounded, taste-gated review manuscript from an approved lean planning package. Uses references.bib, reference-map.json, review-plan.md, evidence-ledger.md, and table-figure-plan.md; use only after plan is approved.
Understand the autor codebase at a high level. Use when the user asks how to use autor, what built-in skills it has, what features it supports, which workflow to choose, or wants a project overview before using other skills.
Write an exploratory literature review based on papers in an autor workspace. For approved review-article workflows, respect the canonical references.bib / reference-map.json / evidence-ledger contract produced by plan.
High-level orchestrator for iterative review planning and hypothesis evolution. Builds the current autor planning package, triggers targeted acquisition, maintains corpus layers, and stops before writing.