agentleFS
Sign inStart free
WUBING2023WUBING2023/PaperSpine/CLAUDE.md

PaperSpine

CLAUDE.md6k starsChanged 41 days agoMITView on GitHub

PaperSpine is a contribution-first paper and report writing skill for Claude Code, Codex, OpenClaw, and Hermes CLI. It researches the target scene, confirms the controlling contribution and motivation with the user, then writes/rewrites/ builds end to end with LaTeX/PDF/Word output. The methodology is enforced by gates: Contribution-First (confirmed_contribution.md), Results-as- Validation (results_validation.md), Reviewer-Aware (reviewer_audit.md). Motivation is still required, but it supports the confirmed contribution. Version: 4.0.0 (canonical: dist/paperspine_version.json, auto-propagated to .claude-plugin/plugin.json and marketplace.json by the sync script). V4 collapsed the old…

We split this CLAUDE.md into the rules it gives an agent

2 rules. About 0 are common, 1 is distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Write LaTeX/Markdown templates as literal file content, not built up from pieces in Python.

    line 60
  2. Very specific:

    Do not push to GitHub unless explicitly asked.

    line 74

≈ “Common” is an estimate: at least 1 in 5 of the 25,680 CLAUDE.md files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
CLAUDE.md

PaperSpine Project Memory (V4)

Project Purpose

PaperSpine is a contribution-first paper and report writing skill for Claude Code, Codex, OpenClaw, and Hermes CLI. It researches the target scene, confirms the controlling contribution and motivation with the user, then writes/rewrites/ builds end to end with LaTeX/PDF/Word output. The methodology is enforced by gates: Contribution-First (confirmed_contribution.md), Results-as- Validation (results_validation.md), Reviewer-Aware (reviewer_audit.md). Motivation is still required, but it supports the confirmed contribution.

Version: 4.0.0 (canonical: dist/paperspine_version.json, auto-propagated to .claude-plugin/plugin.json and marketplace.json by the sync script).

Architecture: ONE skill

V4 collapsed the old 12-skill flat suite into a single paper-spine skill — an orchestrator with 12-stage routing inside SKILL.md, references/ playbooks for each stage, and agents/ role cards for parallel sub-agents. There are no longer separate worker skills.

Single source of truth: src/

  • src/skill/SKILL.md — orchestrator + command/stage routing tables
  • src/skill/references/*.md — per-stage playbooks (intake, research, citation, rewrite, build, humanize, latex, audit, translate, submission, respond, resume, contribution, results-validation, reviewer-audit, scenarios, …)
  • src/skill/agents/* — role cards for fan-out agents (research scene/sota/ exemplar, reviewer method/evidence/writing) + openai.yaml
  • src/scripts/*.py — deterministic gate/check scripts (standard library only): sync_local_installs.py, contribution_check.py, results_validation_check.py, reviewer_audit_check.py, progress_check.py, submission_check.py, respond_check.py, citation_verification_en.py, latex_guard.py, humanize_check.py, style_metrics.py, etc.
  • src/adapters/{claude,codex,hermes} — per-host frontmatter/manifest overlays

Generated: dist/ (never hand-edit)

dist/ is fanned out from src/ by sync_local_installs.py for 4 hosts:

  • dist/claude/skills/paper-spine (+ /paperspine command)
  • dist/codex/skills/paper-spine (+ /paperspine prompt)
  • dist/openclaw/skills/paper-spine
  • dist/hermes/skills/academic-writing/paper-spine

Install entry points: python install.ps1 (Windows) / bash install.sh (macOS/Linux) — thin wrappers delegating to src/scripts/sync_local_installs.py; PS5.1-safe; they never touch settings.json.

Acceptance harness: examples/v4_acceptance.py.

DISCIPLINE — rules guarded by tests

  1. Single source of truth. Edit src/ only; never hand-edit dist/ copies. Regenerate with python src/scripts/sync_local_installs.py --dist-only (project-internal) or --clean-legacy (full sync to install dirs). A dist-sync guard test fails if dist/ drifts from src/.
  2. Never author backslash/LaTeX markdown via Python string concatenation. Write LaTeX/Markdown templates as literal file content, not built up from pieces in Python. A control-char corruption test guards this.
  3. Home isolation. Any code touching ~/.claude / ~/.codex / hermes must take directory arguments (install-dir overrides) and must never touch settings.json. Guarded by validate_repo / install tests.
  4. Standard library only for src/scripts/ Python.

Before claiming ready

python -m pytest tests          # expect 197 passed
ruff check src/scripts tests

Do not push to GitHub unless explicitly asked. Keep README.md (Chinese) and README.en.md (English) content-equivalent.

Download now (ZIP)

Trending

More talk lately.thin dataabout the whole repository

Dec 2025: 0 mentionsJan 2026: 0 mentionsFeb 2026: 0 mentionsMar 2026: 0 mentionsApr 2026: 0 mentionsMay 2026: 0 mentionsJun 2026: 0 mentionsJul 2026: 0 mentionsAug 2026: 0 mentionsSep 2026: 0 mentionsOct 2026: 3 mentionsNov 2026: 0 mentions
Dec 2025Jun 2026Nov 2026
Mentions a month of the WUBING2023/PaperSpine repository that our weekly crawl found: only 3 in all, so this is a hint, not a trend. There is no adoption data for this file yet.

Who's talking about WUBING2023/PaperSpine

The biggest 3 of 3 mentions linking to the repository that our crawl found. Each line is their own headline.

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free