Borrowing it
Nothing to install: this file belongs to ZimoLiao/scholaraio. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/ZimoLiao/scholaraio/main/.claude/skills/search/SKILL.mdgit clone --depth 1 https://github.com/ZimoLiao/scholaraioWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/zimoliao/scholaraio/search)<a href="https://agentmods.dev/skills/zimoliao/scholaraio/search"><img src="https://agentmods.dev/badge/skills/zimoliao/scholaraio/search/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/zimoliao/scholaraio/search"><img src="https://agentmods.dev/badge/skills/zimoliao/scholaraio/search.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00044 | $0.02096 |
| Opus 5 | $0.00022 | $0.01048 |
| Sonnet 5 | $0.00009 | $0.00419 |
| Haiku 4.5 | $0.00004 | $0.00210 |
Grade A, and why
search scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 10d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 141 lines — stays where its author put it; the contents beside it link to each section on GitHub.
文献搜索
在本地论文库中搜索文献。默认使用融合检索(关键词 + 语义向量合并排序),也支持单独使用某一种模式。
执行逻辑
-
解析用户输入,判断搜索模式:
- 如果用户明确要求"语义搜索"、"向量搜索"或"vsearch",使用
vsearch - 如果用户明确要求"关键词搜索"、"全文搜索"或"FTS",使用
search - 如果用户明确要找“证据片段”、“原文片段”、“行号定位”、“在哪一节/哪几行提到”,使用
search --chunk;若 chunk 索引尚未建立,先运行index --chunks - 如果用户明确按作者搜索(如"找某某的论文"、"某某发表的"),使用
search-author - 如果用户要求按引用量排序(如"引用最高的"、"最经典的"、"top cited"),转交
/citationsskill - 默认使用
usearch(融合检索)——同时执行 FTS5 关键词搜索和 FAISS 语义搜索,合并去重排序。两路都命中的论文排名靠前。向量索引不可用时自动降级为纯关键词。 - 如果用户要求跨库搜索(如"也搜一下 arXiv"、"在 explore 库里也找找"、"也搜 proceedings"、"全部来源"、"联邦搜索"),使用
fsearch
- 如果用户明确要求"语义搜索"、"向量搜索"或"vsearch",使用
-
从用户输入中提取:
- 查询词:用户想搜索的内容
- 返回数量:使用规范参数
--limit N;未指定则使用默认值 - 年份过滤:
--year 2023(单年)、--year 2020-2024(范围)、--year 2020-(起始年至今) - 期刊过滤:
--journal "Fluid Mechanics"(模糊匹配) - 类型过滤:
--type review(模糊匹配,常见值:review、journal-article、book-chapter)
查询词拆分原则:不要把“作者 + 年份 + 关键词/题名词”全部拼进同一个 query。这条规则同时适用于
search、vsearch和usearch:search会把整串文本交给 FTS5MATCH;作者缩写、全名、标点或年份 token 只要和索引不一致,就可能让原本可命中的论文搜不出来。vsearch通常不会因此空结果,但作者/年份/期刊等限定词会作为噪声进入 query embedding,可能拉低相关论文分数或引入相近但不精确的结果。usearch同时跑 FTS 和向量;脏 query 可能让 FTS leg 失效,只剩语义命中,结果不再获得both加分。- 年份必须优先放到
--year,不要放进 query。 - 明确按作者找时用
search-author "<作者姓或姓名>",不要把作者混进主题 query。 - 已知题名或主题时,query 保持为最稳定的题名/主题关键词;需要作者/年份约束时分步过滤或二次确认。
- 如果第一次无结果,先去掉作者缩写、年份、机构、期刊等限定词,只保留题名核心词或主题词再搜。
-
执行搜索命令:
融合检索(默认):
scholaraio usearch "<查询词>" --limit <N> [--year <Y>] [--journal <J>] [--type <T>]
关键词搜索:
scholaraio search "<查询词>" --limit <N> [--year <Y>] [--journal <J>] [--type <T>]
证据片段搜索(返回 paper、section、line range、snippet):
scholaraio search --chunk "<查询词>" --limit <N> [--year <Y>] [--journal <J>] [--type <T>]
语义搜索:
scholaraio vsearch "<查询词>" --limit <N> [--year <Y>] [--journal <J>] [--type <T>]
作者搜索:
scholaraio search-author "<作者名>" --limit <N> [--year <Y>] [--journal <J>] [--type <T>]
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 10d ago First seen · 141 lines · 44 tokens per session scan A bc23e4979957
search is a skill published in the GitHub repository ZimoLiao/scholaraio (570 stars, last pushed 10d ago), licensed MIT. It adds 44 tokens to every session and 2,096 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
research-workflow
This skill should be used when the user asks research questions, needs information lookup, wants comparisons, asks "what is", "how does", "explain", "compare", "best practices", "latest developments", or any query requiring web search, documentation lookup, or synthesis of multiple sources. Provides optimal routing…
thoughtdag
Open the current Codex session in ThoughtDAG (desktop app direct; local bridge fallback — data never leaves this machine), or query local session history with why / find / recall.
gigaxity-deep-research
Deep research MCP server wrapping Qwen3-30B-A3B-Thinking via OpenRouter. Use when an agent needs cross-source synthesis with citations, exploratory expansion of an unfamiliar topic, chain-of-thought reasoning over evidence, or fast conversational lookups grounded in live web search. Exposes six MCP tools — two…
template-agentic-use
Agentic-use and skill-routing workflow for the Research Project Template. USE WHEN discovering relevant skills, making template easier for agents to navigate, auditing docs/prompts skill coverage, checking .cursor/skillmanifest.json, evaluating external skills, reviewing external agentic operating models such as…
infrastructure-overview
Top-level skill for the research template infrastructure layer. Use in Cursor, Claude Code, or similar agents when editing or importing anything under infrastructure/, understanding the two-layer architecture, or wiring build/validation/rendering/publishing. Covers module discovery, import patterns, thin…
infrastructure-search
Discovery utilities. Hosts subpackages — literature for Paperclip-style multi-source academic search; exa for Exa-backed web search, contents, answers, and find-similar; monid for the Monid gateway (discover/inspect/run) plus offline USD-per-1k search API pricing; and deepresearch for provider-neutral long-running…