Is the Literature Review Agent skill free?

Yes. The skill itself is free from Ar9av/PaperOrchestra. SkillProof publishes the install command and an independent test verdict at no cost.

What is the Literature Review Agent SkillProof Score?

8.4/10 — installs cleanly 5/5, triggers reliably 5/5, output vs. baseline 7/10, docs & honesty 4/5.

How do I install Literature Review Agent?

Copy the install command from this page, run it in your terminal, and restart Claude Code. Skills live in ~/.claude/skills/ (global) or .claude/skills/ inside a project.

Can I use Literature Review Agent with Cursor, Copilot, Gemini CLI, Codex or other AI tools?

The SKILL.md format is native to Claude (Claude Code, Desktop, claude.ai). The instructions inside adapt to other assistants: Cursor rules, GitHub Copilot instructions, Windsurf rules, Custom GPTs, AGENTS.md for OpenAI Codex, and GEMINI.md for Google Gemini CLI — our conversion guides cover each, and the free converter on the tools page does the wrapping for you.

Literature Review Agent

Verifies paper citations, flags hallucinated ones, builds BibTeX, drafts related work

by Ar9av · Ar9av/PaperOrchestra

Tested · Works

Install — copy the command Download source ↗

Test report

Verdict: Tested · Works
Score: 8.4/10
Tested: Jul 21, 2026
Environment: Claude Code 2.x (agent harness)
Upstream re-checked: Aug 10, 2026 · 0559a83

⚠ The author changed this skill after we tested it. The verdict below describes the version we ran on Jul 21, 2026; it's queued for a re-test.

Cloned the repo; all 14 scripts and every spot-checked reference file (bibtex_format.py, levenshtein_match.py, dedupe_by_id.py, references/prompt.md) return HTTP 200. Ran bibtex_format.py on a hand-built 3-paper pool: it emitted deterministic stopword-aware keys (vaswani2017attention, zhao2023survey, he2016deep), wrote eprint/doi only when present, and wrote bibtex_key back into citation_pool.json — the baseline hand-written bib would use ad-hoc keys and never populate the pool, breaking the downstream coverage gate. check_cutoff.py and citation_coverage.py behaved as specified (coverage gates on a floor(0.9*N) count, so 2/3 'passed' at count 2 — slightly looser than the SKILL's '≥90% of pool' wording). levenshtein_match.py fails with a clear 'pip install python-Levenshtein' error until that undocumented dep is installed, after which it returned 100 PASS / 31 FAIL correctly. Web-discovery + live Semantic Scholar/Crossref/OpenAlex verification were not exercised (need host web search + network), so the output score reflects only the deterministic BibTeX/coverage slice. No security smells: keys read from env, no curl|sh, no exfiltration.

Scored on four weighted criteria — install, triggering, output vs. baseline, docs. How scoring works

Installs cleanly 5/5
Triggers reliably 5/5
Output vs. baseline 7/10
Docs & honesty 4/5

What Literature Review Agent does

Runs the citation stage of an academic-paper pipeline: discovers candidate papers via web search, verifies each through Semantic Scholar (fuzzy title match, temporal cutoff, dedup), cross-checks against Crossref and OpenAlex to flag hallucinated references, builds a BibTeX file with deterministic keys, and drafts the Introduction and Related Work sections. Triggers when asked to find citations for a paper, draft the related work, or build a verified bibliography.

How to install Literature Review Agent

git clone --depth 1 https://github.com/Ar9av/PaperOrchestra.git /tmp/literature-review-agent-src
mkdir -p ~/.claude/skills
cp -R /tmp/literature-review-agent-src/skills/literature-review-agent ~/.claude/skills/literature-review-agent
# Undocumented runtime dep: scripts/levenshtein_match.py needs python-Levenshtein
pip install python-Levenshtein
# Optional deps used only if you enable those backends / cross-index checks:
#   requests            (s2_search.py, crossref_client.py, openalex_client.py, exa_search.py)
# Optional API keys (all read from env, none committed):
#   SEMANTIC_SCHOLAR_API_KEY  — higher S2 rate limit (public endpoint works without it)
#   EXA_API_KEY               — optional Exa web-search backend for Phase 1
#   PAPER_ORCHESTRA_MAILTO    — polite-pool email for Crossref/OpenAlex
# Designed as Step 3 of the PaperOrchestra pipeline; needs a workspace/outline.json + host web search for the full run.

Skills live in ~/.claude/skills/ (global) or .claude/skills/ (per-project). Restart Claude Code after installing.

Commands — how to trigger Literature Review Agent

/literature-review-agent Verifies paper citations, flags hallucinated ones, builds BibTeX, drafts related work

It also activates on plain-language prompts like these:

Find citations for my paper's related work section
Draft the related work from this outline.json
Verify these citations against Semantic Scholar

Frequently asked questions

Is the Literature Review Agent skill free?: Yes. The skill itself is free from Ar9av/PaperOrchestra. SkillProof publishes the install command and an independent test verdict at no cost.
Does Literature Review Agent work with Claude Code?: We tested it with Claude Code 2.x (agent harness) on Jul 21, 2026. Verdict: Tested · Works. Cloned the repo; all 14 scripts and every spot-checked reference file (bibtex_format.py, levenshtein_match.py, dedupe_by_id.py, references/prompt.md) return HTTP 200. Ran bibtex_format.py on a hand-built 3-paper pool: it emitted deterministic stopword-aware keys (vaswani2017attention, zhao2023survey, he2016deep), wrote eprint/doi only when present, and wrote bibtex_key back into citation_pool.json — the baseline hand-written bib would use ad-hoc keys and never populate the pool, breaking the downstream coverage gate. check_cutoff.py and citation_coverage.py behaved as specified (coverage gates on a floor(0.9*N) count, so 2/3 'passed' at count 2 — slightly looser than the SKILL's '≥90% of pool' wording). levenshtein_match.py fails with a clear 'pip install python-Levenshtein' error until that undocumented dep is installed, after which it returned 100 PASS / 31 FAIL correctly. Web-discovery + live Semantic Scholar/Crossref/OpenAlex verification were not exercised (need host web search + network), so the output score reflects only the deterministic BibTeX/coverage slice. No security smells: keys read from env, no curl|sh, no exfiltration.
What is the Literature Review Agent SkillProof Score?: 8.4/10 — installs cleanly 5/5, triggers reliably 5/5, output vs. baseline 7/10, docs & honesty 4/5.
How do I install Literature Review Agent?: Copy the install command from this page, run it in your terminal, and restart Claude Code. Skills live in ~/.claude/skills/ (global) or .claude/skills/ inside a project.
Can I use Literature Review Agent with Cursor, Copilot, Gemini CLI, Codex or other AI tools?: The SKILL.md format is native to Claude (Claude Code, Desktop, claude.ai). The instructions inside adapt to other assistants: Cursor rules, GitHub Copilot instructions, Windsurf rules, Custom GPTs, AGENTS.md for OpenAI Codex, and GEMINI.md for Google Gemini CLI — our conversion guides cover each, and the free converter on the tools page does the wrapping for you.