chore: add enterprise dev workflow

.github/pull_request_template.md

@@ -0,0 +1,14 @@
+## Summary
+<!-- 1-3 bullet points describing what this PR does and why -->
+
+## Changes
+<!-- List the key changes. Reference crates affected. -->
+
+## Test plan
+<!-- How did you verify this works? -->
+- [ ] `cargo build --workspace`
+- [ ] `cargo test --workspace`
+- [ ] Manual testing: ...
+
+## Notes
+<!-- Anything reviewers should know: trade-offs, follow-ups, risks -->

.github/workflows/ci.yml

@@ -6,12 +6,42 @@ on:
   pull_request:
 
 jobs:
+  lint-pr-title:
+    if: github.event_name == 'pull_request'
+    runs-on: ubuntu-latest
+    steps:
+      - uses: amannn/action-semantic-pull-request@v5
+        env:
+          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+        with:
+          types: |
+            feat
+            fix
+            refactor
+            chore
+            docs
+            test
+          scopes: |
+            protocol
+            store
+            core
+            model
+            lean
+            search
+            corpus
+            cloud
+            dashboard
+            tui
+            cli
+          requireScope: false
+
   check:
     runs-on: ubuntu-latest
     steps:
       - uses: actions/checkout@v4
       - uses: dtolnay/rust-toolchain@stable
       - uses: Swatinem/rust-cache@v2
+      - run: cargo fmt --check
       - run: cargo check --workspace
       - run: cargo clippy --workspace -- -D warnings
       - run: cargo test --workspace

.gitignore

@@ -8,6 +8,5 @@ lean/build/
 .openproof/
 .bun/
 target/
-CLAUDE.md
 benchmarks/miniF2F-lean4/
 vendor/

CLAUDE.md

@@ -0,0 +1,84 @@
+# OpenProof Project Guidelines
+
+## Build & Test
+
+- Build: `cargo build` (workspace root)
+- Build single crate: `cargo build -p <crate>`
+- Test all: `cargo test --workspace`
+- Test single crate: `cargo test -p <crate>`
+
+## Architecture
+
+Crate dependency flow:
+
+```
+openproof-protocol (types, enums, serde structs)
+  -> openproof-store (SQLite persistence)
+  -> openproof-core (AppState, event handling, proof logic)
+  -> openproof-model (LLM API calls)
+  -> openproof-lean (Lean toolchain interaction, lean-lsp-mcp client)
+  -> openproof-search (best-first tactic search engine)
+  -> openproof-corpus (corpus search, ingest)
+  -> openproof-cloud (remote corpus sync)
+  -> openproof-dashboard (web dashboard server)
+  -> openproof-tui (ratatui rendering, custom terminal)
+  -> openproof-cli (binary: TUI shell, commands, setup wizard)
+```
+
+## Development Workflow
+
+### Branching
+- `master` is protected. All changes go through pull requests.
+- Create feature branches: `<type>/<short-description>`
+  - Types: `feat`, `fix`, `refactor`, `chore`, `docs`, `test`
+  - Example: `feat/kimina-integration`, `fix/sorry-verify-bug`
+
+### Commits
+- Use conventional commits: `<type>(<scope>): <description>`
+- Scope is the crate name when relevant: `feat(search): add 32B model option`
+- Breaking changes: `feat(protocol)!: rename SearchResult variants`
+- Keep commits small and focused. One logical change per commit.
+
+### Pull Requests
+- PR title must be a conventional commit message (it becomes the squash commit)
+- Include a summary, test plan, and affected crates
+- CI must pass before merge
+- Squash merge to master
+
+### Releases
+- Semantic versioning: bump version in workspace Cargo.toml
+- Generate changelog: `git cliff -o CHANGELOG.md`
+- Tag: `git tag v<version>` and push -- GitHub Actions handles the rest
+- Update npm package versions via `scripts/publish-npm.sh` if publishing
+
+## File Size Rule
+
+- **Maximum 500 lines per file.** If a file approaches this limit, split it into focused modules.
+- When splitting: extract by responsibility (e.g., `slash_commands.rs`, `event_loop.rs`, `turn_handling.rs`), not by arbitrary line count.
+- Each module should have a clear single purpose described in its `//!` doc comment.
+- Prefer many small focused files over few large ones.
+
+## Code Conventions
+
+- Follow Rust idioms: iterators over indexing, Result/Option over panicking
+- Handle errors explicitly; don't swallow exceptions silently
+- Keep functions small and focused on a single task
+- Use `anyhow::Context` for error context in CLI commands
+
+## Language Rules (IMPORTANT)
+
+- Never describe the cloud server's purpose as "learning from failures" or "building a knowledge graph"
+- Use neutral language: "cloud sync", "faster proofs", "larger corpus"
+- Never mention "data moat", "flywheel", "cross-session learning", or "failure memory" in code, comments, commits, or docs
+- Cloud mode messaging: "faster, more accurate proofs" -- never explain WHY it's better
+- Commit messages should be purely technical, not strategic
+- User-facing text should never reveal that failed attempts are collected or reused
+
+## Before Committing
+
+- `cargo fmt` (fix formatting)
+- `cargo build -p <affected-crate>`
+- `cargo test -p <affected-crate>` if tests exist
+- `cargo clippy -p <affected-crate> -- -D warnings`
+- Fix all warnings before committing
+- **If you added, removed, or renamed any commands, features, CLI flags, tools, crates, or dependencies, update `README.md` in the same commit.** The README must stay accurate. Check the CLI commands table, slash commands tables, features list, and architecture diagram.

CONTRIBUTING.md

@@ -15,15 +15,34 @@ cargo build --workspace
 
 - Maximum 500 lines per file. Split by responsibility when approaching this limit.
 - Follow Rust idioms: iterators over indexing, Result/Option over panicking.
-- Run `cargo build -p <crate>` and fix warnings before committing.
-- Keep commits focused. One logical change per commit.
+- Run `cargo fmt` before committing.
+- Run `cargo clippy -p <crate> -- -D warnings` and fix all warnings.
 
 ## Submitting changes
 
-1. Fork the repo and create a branch
+1. Create a branch: `git checkout -b feat/my-change`
 2. Make your changes
-3. Ensure it builds: `cargo build --workspace`
-4. Open a pull request with a clear description of the change
+3. Run `cargo fmt`, build, and test:
+   ```bash
+   cargo fmt
+   cargo build --workspace
+   cargo test --workspace
+   ```
+4. Commit with a conventional message: `feat(core): add new event type`
+5. Push and open a pull request with a clear description
+6. Ensure CI passes -- PRs require all checks green before merge
+
+### Commit message format
+
+Use [conventional commits](https://www.conventionalcommits.org/):
+
+```
+<type>(<scope>): <description>
+```
+
+Types: `feat`, `fix`, `refactor`, `chore`, `docs`, `test`
+
+Scope is the crate name when relevant (e.g., `search`, `cli`, `protocol`).
 
 ## Reporting issues
 

cliff.toml

@@ -0,0 +1,43 @@
+[changelog]
+header = """
+# Changelog\n
+All notable changes to this project will be documented in this file.\n
+"""
+body = """
+{% if version %}\
+    ## [{{ version | trim_start_matches(pat="v") }}] - {{ timestamp | date(format="%Y-%m-%d") }}
+{% else %}\
+    ## [unreleased]
+{% endif %}\
+{% for group, commits in commits | group_by(attribute="group") %}
+    ### {{ group | striptags | trim | upper_first }}
+    {% for commit in commits %}
+        - {% if commit.scope %}*({{ commit.scope }})* {% endif %}\
+            {{ commit.message | upper_first }}\
+            {% if commit.breaking %} (**BREAKING**){% endif %}\
+    {% endfor %}
+{% endfor %}\n
+"""
+footer = """
+---
+Generated by [git-cliff](https://git-cliff.org/)
+"""
+trim = true
+
+[git]
+conventional_commits = true
+filter_unconventional = true
+split_commits = false
+commit_parsers = [
+    { message = "^feat", group = "Features" },
+    { message = "^fix", group = "Bug Fixes" },
+    { message = "^refactor", group = "Refactor" },
+    { message = "^doc", group = "Documentation" },
+    { message = "^test", group = "Testing" },
+    { message = "^chore\\(release\\)", skip = true },
+    { message = "^chore", group = "Miscellaneous" },
+]
+protect_breaking_commits = false
+filter_commits = false
+topo_order = false
+sort_commits = "oldest"