菜单
Search academic papers in the local autor knowledge base. Uses auditable node-level FTS5 evidence search, evidence bundles, author search, and top-cited ranking. Use when the user wants to find papers, look up literature, search by author, or explore research topics.
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 | search |
| description | Search academic papers in the local autor knowledge base. Uses auditable node-level FTS5 evidence search, evidence bundles, author search, and top-cited ranking. Use when the user wants to find papers, look up literature, search by author, or explore research topics. |
Search for papers in the local library. Defaults to auditable node-level FTS5 retrieval over metadata plus paper.md chunks. For complex questions, generate a bundle with trace/verify artifacts.
Parse user input and determine the search mode:
search or researchsearchresearchsearch-authortop-citedsearch for paper retrieval, or research when answer-time provenance is needed.Extract from user input:
--top N if specified; otherwise use the default--year 2023 (single year), --year 2020-2024 (range), --year 2020- (from year onwards)--journal "Fluid Mechanics" (fuzzy match)--type review (fuzzy match; common values: review, journal-article, book-chapter)Run the search command:
Auditable search (default):
autor search "$ARGUMENTS" --top <N> [--year <Y>] [--journal <J>] [--type <T>]
Keyword search:
autor search "<query>" --top <N> [--year <Y>] [--journal <J>] [--type <T>]
Evidence bundle:
autor research "<query>" --top <N> --run-dir workspace/runs/<case>
Author search:
autor search-author "<author name>" --top <N> [--year <Y>] [--journal <J>] [--type <T>]
Top-cited ranking:
autor top-cited --top <N> [--year <Y>] [--journal <J>] [--type <T>]
Present the search results to the user. Each result may include evidence snippets with node_id, section, snippet, and ref_path. For research, answer only from the generated bundle and preserve the ## References section.
Complex queries: when CLI parameter combinations are insufficient (e.g. filtering by first-author initial, multi-condition intersections, custom sorting), read data/papers/*/meta.json directly in Python. Key JSON fields:
title, authors, first_author, first_author_lastname, year, doi, journal,
abstract, paper_type, citation_count (dict: crossref/semantic_scholar/openalex),
ids, toc, l3
User says: "Search for papers on turbulent boundary layers."
→ Run search "turbulent boundary layer"
User says: "Use semantic search to find literature on drag reduction; give me the top 5."
→ Say vector search is deprecated, then run search "drag reduction" --top 5
User says: "Find papers by Liao Z-M."
→ Run search-author "Liao"
User says: "What are the most-cited papers in my library?"
→ Run top-cited --top 10
User says: "Papers on drag reduction after 2020."
→ Run search "drag reduction" --year 2020-
User says: "Turbulence papers published in JFM."
→ Run search "turbulence" --journal "Fluid Mechanics"
User says: "Most-cited review articles in my library."
→ Run top-cited --top 10 --type review
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.