Literature Review Agent (PaperOrchestra Step 3)

Discovers candidate papers via web search, verifies them through Semantic Scholar plus Crossref/OpenAlex cross-checks, and drafts a cited Introduction + Related Work with a matching refs.bib.

EducationAdvanced★ 644⑂ 91AI score 9/10Last updated: Aug 9, 2026

What it does

Implements Step 3 of the PaperOrchestra paper-writing pipeline.

Phase 1 — Discovery: runs the search directions and Related Work cluster queries from outline.json through your host's web search (WebSearch, or optional Exa / Tavily backends) to collect candidate papers.
Pre-dedup: pre_dedup_candidates.py removes duplicate candidates up front, saving 30–40% of the Semantic Scholar quota.
Phase 2 — Verification: sequential, cached, 1 QPS S2 lookups with Levenshtein title match > 70, non-empty abstract requirement, and a venue-derived temporal cutoff; then dedup by paperId.
Cross-index check: re-verifies every paper against Crossref and OpenAlex and grades each as high / medium / low / conflict to surface hallucinated citations (a WARN gate, not a hard block).
Outputs: citation_pool.json, refs.bib, and intro_relwork.tex citing ≥90% of the verified pool, enforced by a coverage gate.

Who it's for

Researchers and grad students writing to a conference deadline (CVPR, ICLR, NeurIPS)
Anyone burned by LLM-fabricated references
Teams automating bibliography construction and Related Work drafting
Users running the full PaperOrchestra pipeline

Examples

"Find citations for my paper" → executes discovery queries and produces a verified pool plus refs.bib.
"Draft the related work section" → writes LaTeX Introduction + Related Work citing at least 90% of the pool, with timeline and evaluation guardrails.
"Check my bibliography for hallucinated citations" → runs cross_verify.py and returns a tiered corroboration report.

· · · Install guide · · ·

Try it now, no install

Paste this into Claude to use the skill without installing anything.

Read the instructions in this file and follow them to help me:
https://raw.githubusercontent.com/Ar9av/PaperOrchestra/HEAD/skills/literature-review-agent/SKILL.md

What I want: (describe your task here)

If Claude can't open the link, open it yourself and paste the contents instead.

↓ If it works for you, download the ZIP below and install it. Then it runs on its own — no pasting each time.

Install in the Claude app (no terminal)

Download the ZIP with the button below.
In Claude, open Settings → Capabilities and turn on 'Code execution and file creation'. (one time)
Go to Customize → Skills → + → 'Upload a skill' and upload the ZIP.

↓ Download ZIP

Install in Claude Code

Let Claude do it — paste this into Claude Code

Install the skill I found on Claude Skill Mart.
Copy the skills/literature-review-agent folder from the GitHub repo Ar9av/PaperOrchestra into my ~/.claude/skills/literature-review-agent/.
When it's done, tell me in one line what this skill can do.

Install with a command instead

git clone https://github.com/Ar9av/PaperOrchestra.git /tmp/PaperOrchestra && mkdir -p ~/.claude/skills && cp -r /tmp/PaperOrchestra/skills/literature-review-agent ~/.claude/skills/

⚠ This is a third-party skill. Check the source repository before installing.

Open a terminal.
Clone the repo: git clone https://github.com/Ar9av/PaperOrchestra.git
Create the skills directory: mkdir -p ~/.claude/skills
Copy the skill: cp -r PaperOrchestra/skills/literature-review-agent ~/.claude/skills/
Recommended — also copy the orchestrator and shared resources it references: cp -r PaperOrchestra/skills/paper-orchestra PaperOrchestra/skills/shared ~/.claude/skills/
Make sure Python 3 and the script dependencies (e.g. requests) are installed.
Optional API keys: export SEMANTIC_SCHOLAR_API_KEY=..., export EXA_API_KEY=... or export TAVILY_API_KEY=..., and export PAPER_ORCHESTRA_MAILTO=you@example.com
Restart Claude Code and ask: "Find citations for my paper and draft the related work."

View source on GitHub ↗License: NOASSERTION