Add code-hygiene checks (large-file, folder-size, AI-writing) across prek + CI + cloud

[Miyamura80]

Ports four code-hygiene checks into this repo, each wired at three layers kept in sync (prek, GitHub Actions, Claude Code cloud).

Checks

1. Large-file check — fail when a source file exceeds a line-count threshold (warn 500 / error 800).
2. Folder-size check — fail when a folder has too many immediate source children (warn 20 / error 35).
3. Em-dash checker — fail on — (U+2014) anywhere in the repo.
4. Contrastive-parallelism checker — fail on the "not just X, but Y" AI-writing tell.

This repo is Rust, so the rs stack was used throughout: SOURCE_EXTS=(rs), prek files = \.(rs)$, CI paths: '**.rs' — all the same extension set.

Files

scripts/
  check_large_files.sh      line-count guard (SOURCE_EXTS=rs)
  check_folder_sizes.sh     folder-fanout guard (SOURCE_EXTS=rs)
  check_ai_writing.sh       em-dash + contrastive parallelism (ripgrep, git-grep fallback; no build)
prek.toml                   upstream check-added-large-files + 3 local hooks
.github/workflows/
  large-files.yaml          Large File Check
  folder-size.yaml          Folder Size Check
  ci.yml                    + AI writing check step
.claude/
  settings.json             SessionStart -> hooks/session-start.sh
  hooks/session-start.sh    cargo fetch + prek install (remote-session guarded)

Only the Rust AI-writing impl exists (no .py/.ts variants).

Notes

• Replaced 27 pre-existing em dashes in ARCHITECTURE.md, CONTRIBUTING.md, README.md, SECURITY.md with hyphens so the em-dash check passes on a clean tree.
• No contrastive-parallelism hits in the existing tree, so the full high-confidence pattern set is kept (including goes beyond / so much as).

Verification (all exit 0)

• check_large_files.sh --all — clean; daemon.rs (645) and env_store.rs (555) warn but stay under the 800 error threshold.
• check_folder_sizes.sh --all — clean.
• check_ai_writing.sh — passes on both the rg and git grep paths.
• prek run --all-files — the three local hooks pass.

🤖 Generated with Claude Code

---

Generated by Claude Code

---

Summary by cubic

Add repo-wide hygiene checks for large Rust sources (file length, folder fan-out, and writing-style patterns) running in local prek, GitHub Actions, and cloud editor sessions. Warn/error thresholds: 500/800 lines per file and 20/35 files per folder.

• New Features

  • Shell checks for large files, folder size, and writing style scoped to *.rs.
  • GitHub Actions: new Large File and Folder Size workflows; AI-writing step in ci.yml.
  • Local and cloud: prek.toml hooks; .claude SessionStart installs Rust, sets PATH, installs prek, runs cargo fetch.

• Bug Fixes

  • Docs merge with main: adopted the rewritten Components section and README badges; re-applied the hyphen sweep across docs and docs/architecture.svg to pass the writing check.
  • Cloud: ensure ~/.cargo/bin is on PATH and source ~/.cargo/env after a fresh install so cargo fetch and checks run reliably.

^{Written for commit 638b7a0. Summary will update on new commits.}

[qodo-code-review]

Qodo reviews are paused for this user.

Troubleshooting steps vary by plan Learn more →

On a Teams plan?
Reviews resume once this user has a paid seat and their Git account is linked in Qodo.
Link Git account →

Using GitHub Enterprise Server, GitLab Self-Managed, or Bitbucket Data Center?
These require an Enterprise plan - Contact us
Contact us →

[cubic-dev-ai]

1 issue found across 13 files

<sub>Reply with feedback, questions, or to request a fix.

Re-trigger cubic</sub>