0xmariowu/autosearch/papers

Skill Attribution

Source: self-written for task F203.

name: papers description: Multi-source academic paper search (arxiv, pubmed, biorxiv, medrxiv, google_scholar) via paper-search-mcp. version: 1 languages: [en, mixed] methods:

• id: via_paper_search

impl: methods/via_paper_search.py requires: [] rate_limit: {per_min: 20, per_hour: 300} fallback_chain: [via_paper_search] when_to_use: query_languages: [en, mixed] query_types: [academic-paper, research, technical-deep-dive] domain_hints: [science, medicine, biology, cs, physics, chem] quality_hint: typical_yield: high chinese_native: false layer: leaf domains: [academic] scenarios: [paper-collection, survey] model_tier: Fast experience_digest: experience.md

Overview

papers is the broad academic search channel for research-heavy queries that benefit from cross-checking multiple paper sources instead of relying on a single index. It aggregates arXiv, PubMed, bioRxiv, medRxiv, and Google Scholar so the planner can surface preprints, medical literature, and general scholarly coverage through one paper-oriented channel.

Known Quirks

• Google Scholar can rate-limit or degrade if queried too aggressively, so this channel uses conservative limits.
• bioRxiv can be flaky on some requests and may intermittently return no results even when related sources succeed.
• Overlap with the standalone arxiv channel is intentional; duplicate arXiv fetches are acceptable because reranking dedupes by URL later.

Quality Bar

• Evidence items have non-empty title and url.
• No crash on empty or malformed API response.
• Source channel field matches the channel name.