research-lookup_skill

This skill enables real-time research lookup across Perplexity Sonar models, selecting the optimal probe for complex queries and returning cited academic
  • Python

7.4k

GitHub Stars

1

Bundled Files

2 months ago

Catalog Refreshed

4 months ago

First Indexed

Readme & install

Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.

Installation

Preview and clipboard use veilstrat where the catalogue uses aiagentskills.

npx veilstrat add skill k-dense-ai/claude-scientific-skills --skill research-lookup

  • SKILL.md20.3 KB

Overview

This skill provides real-time research lookups using Perplexity Sonar models via OpenRouter. It automatically chooses between Sonar Pro Search for fast lookups and Sonar Reasoning Pro for deeper analysis, returning summaries, citations, and source links. Responses prioritize peer-reviewed literature, recent studies, and proper bibliographic details.

How this skill works

The skill inspects the user query for complexity indicators and either runs a fast scholarly search (Sonar Pro Search) or a reasoning-enabled synthesis (Sonar Reasoning Pro). It retrieves papers, reports, and technical documents from academic and reputable sources, extracts key findings and metadata, and formats results with complete citations, DOIs or stable URLs, and access dates. A force_model parameter lets you override automatic selection.

When to use it

  • Fetch current papers, reviews, and emerging findings in a field
  • Verify facts, statistics, or methodological details against primary literature
  • Gather citation-ready sources for manuscripts, grants, or reports
  • Compare methods, mechanisms, or results across studies
  • Look up protocols, specifications, or technical procedures
  • Obtain statistical summaries and source-attributed data for analysis

Best practices

  • Frame queries with topic + specific aspect + time frame + type of information (e.g., ‘CRISPR + off-target effects + 2022-2024 + review’)
  • Use reasoning keywords (compare, explain, analyze) if you need synthesis or trade-off evaluation
  • Force a model only when you know you need speed (pro) or deeper synthesis (reasoning)
  • Request full citations or DOI links when preparing manuscripts or reference lists
  • Break very broad queries into focused sub-questions for more precise results
  • Follow up asking for limitations, methodologies, or similar studies to deepen context

Example use cases

  • Literature review: find and cite the latest 3–5 papers on a niche topic (with DOIs)
  • Protocol lookup: retrieve step-by-step methods and critical parameters for laboratory techniques
  • Comparative analysis: synthesize evidence comparing two treatment modalities with cited sources
  • Statistical lookup: obtain recent prevalence or survey statistics with source attribution
  • Mechanism explanation: explain causal relationships across multiple studies with referenced evidence

FAQ

It scores query complexity using reasoning keywords, punctuation and length; a threshold triggers Sonar Reasoning Pro for synthesis, otherwise Sonar Pro Search handles lookups. You can override this with force_model.

Will responses include full citation details?

Yes. Outputs include complete bibliographic info, DOIs or stable URLs, and access dates when available to support manuscript and reporting needs.

Built by
VeilStrat
AI signals for GTM teams
© 2026 VeilStrat. All rights reserved.All systems operational
research-lookup skill by k-dense-ai/claude-scientific-skills | VeilStrat