google-research/paper-orchestra

PaperOrchestra is a multi-agent framework for automated AI research paper writing.

Python★ 1,740+313 that day#11 on trending

Analyzed November 17, 2026 at d9bae99 (main) with jscpd v5 in 0.9s.

E30.2/100
Project health9,722 lines of code (S)
Duplication56
7.9% in python, latex, bash
Dead code7
34%
Complexity62
37.8%

The score is jscpd's own --health model: markup, data and text left out, tests kept. The numbers below are the trending scan's.

7.93%duplicated code · 718 of 9,055 lines
22clones in code
32code files of 45 scanned
7.38%duplicated over all files · 24 clones

The code number leaves out tests, fixtures, docs, examples, data files, styles, translations and vendored code.

Largest clones in the code

  • 117 lines in 2 placespython566 tokens
    253    writer_model_name=writer_model_name,
    254    research_cutoff=research_cutoff,
    255)
    256
    257if success:
    258    result["status"] = "success"
    259    result["paper_path"] = final_paper_path
    260    print(f"Paper {paper_id} successfully written!")
    261else:
    262    print(f"Paper {paper_id} writing failed.")
    

    First 10 of 117 lines.

  • 92 lines in 2 placespython385 tokens
    237with ThreadPoolExecutor(max_workers=max_workers) as executor:
    238    futures = []
    239    for pid, p1, p2 in common_ids:
    240        f = executor.submit(
    241            _process_and_save_single_sxs,
    242            pid,
    243            p1,
    244            p2,
    245            output_json_path,
    246            current_results,
    

    First 10 of 92 lines.

  • 68 lines in 2 placespython323 tokens
    105rating2 = rate_sxs_lit_review(paper2_text, paper1_text, model_name=model_name)
    106
    107if not rating1 or not rating2:
    108    print(
    109        f"❌ [{paper_id}] Failed to get ratings for one or both rounds.", flush=True
    110    )
    111    return paper_id, False
    112
    113winner1 = rating1.get("winner", "").lower() if isinstance(rating1, dict) else ""
    114winner2 = rating2.get("winner", "").lower() if isinstance(rating2, dict) else ""
    

    First 10 of 68 lines.

7 more
Largest clones over all files, tests and data included
By format, code only
FormatFilesLinesClonesDuplicated linesDuplication
python319,032227187.95%
bash123000%
By format, all files
FormatFilesLinesClonesDuplicated linesDuplication
python319,032227187.95%
latex67562486.35%
markdown6569000%
bash123000%
text16000%

Trending appearances

DayRankStarsFilesLinesClonesAll filesCodeHealthCommit
Nov 17, 2026#111,740 (+313)4510,386247.38%7.93%E 30.2d9bae99

Want this for your own project? Install jscpd and run jscpd ., or measure it the way the baselines were measured.