Skip to content

Integration preview: 2.3.0 relaunch (do not merge) - #158

Draft
PatrickSys wants to merge 24 commits into
masterfrom
integration/relaunch-2.3.0
Draft

PatrickSys wants to merge 24 commits into
masterfrom
integration/relaunch-2.3.0

Conversation

@PatrickSys

Copy link
Copy Markdown
Owner

Agent: cursor-cloud/cbc-integration

PR Merge result Tests after Conflicts and fixes
#154 fast-forward pass (131 files, build, pnpm pack) none
#145 merged pass auto-merge docs/capabilities.md only
#142 merged (PR closed, branch merged) pass none
#146 merged pass capabilities.md (OpenAI + Ollama bullets); types.ts (resolveDefaultApiEndpoint, dedupe); embedding env + OpenAI tests
#147 merged pass none
#153 merged pass none
#155 merged pass none
#156 merged pass docs/benchmark.md: kept front-end matrix (#155) + niche benchmark (#156)
#148 merged pass README/cli/capabilities release copy; README/CHANGELOG niche numbers aligned post-#156 (integration-only; #148 branch still needs patch below)

Final integration checks: pnpm test (1348 passed), pnpm run build, pnpm pack --dry-run, MCP stdio search_codebase on a Vite React template (App.tsx query returned 5 hits).

#148 README/CHANGELOG patch (after #156 lands first)

Add to README (near existing 25.7% retrieval line):

The reproducible niche prior-art slice (18 tasks, npm run benchmark:niche) scored 50.0% top-3 file hit rate overall but 0/6 top-3 file hits on team-convention lookup (pattern signals matched; ranked paths did not).

CHANGELOG 2.3.0 Docs bullet — append:

…and the separate niche prior-art run (50.0% top-3 overall, 0/6 team-convention file hits) per docs/benchmark.md…

README Evidence section — state the niche table is a separate frozen family, not pooled with the 25.7% retrieval metric.

Recommended master merge order: #154 → #145 → #142 → #146 → #147 → #153 → #155 → #156 → #148 (with doc patch).

Do not merge this integration PR into master.

Open in Web Open in Cursor 

Init configures Claude Code, Codex, Cursor, and OpenCode together and saves the Node binary plus the package file from the npx cache. map indexes before it prints, --version prints the package version, and a low-confidence search tells the agent a weak result is not proof the code is missing.
Implement OllamaEmbeddingProvider using /api/embed with OLLAMA_HOST
and provider-specific model defaults. Wire getEmbeddingProvider,
dimension detection, and skip Transformers runtime for ollama.
Includes OPENAI_BASE_URL config compatible with PR #145, docs, and
HTTP-mocked tests.
Pin onnxruntime-node 1.24.3 for npm installs, assert a minimum runtime
before Transformers.js embedding/rerank work, and add regression tests.

Fixes #68
Align README and docs with team-conventions plus edit-preflight positioning,
credit Aider/Repomix/jCodeMunch, and separate shipped 2.3.0 from planned 2.4.0.
Add 2.3.0 CHANGELOG release notes including the npm 2.2.0 silent CLI fix.
Configure Claude Code, Cursor, Codex, VS Code, Windsurf, and OpenCode in one
pass with idempotent merges, user-scoped Windsurf paths, and explicit create/
update output. Add temp-HOME tests and document init in the README.
Port canonical realpath entrypoint matching from master so npx-style
launchers named codebase-context dispatch the CLI instead of exiting silently.
Clarify searchable chunk limits, exclude vendored bundles from bounded map,
and boost keyword hits for file-like query tokens. Adds regression tests.
OPENAI_BASE_URL wiring is useless if a cached OpenAI client keeps the first endpoint.
Bring PR #142 up to date with model-free indexing, init preflight, and
master CLI entrypoint detection while keeping symlink --version test and
search-quality hints.
Fix stale first-run model-download wording; fold product-candidate notes
into 2.3.0 (unreleased/TBD); thank @rothnic for #68/#70 analysis.
Add .vue/.svelte to scan and language detection, extract script blocks for
generic analysis, detect Nuxt/SvelteKit from workspace deps, and ignore
SvelteKit + routes and non-Next app/pages paths. Document five-repo smoke
results in docs/benchmark.md with fixture tests.
Extend the existing eval fixtures with a frozen 18-task manifest across
three public repos, a one-command runner comparing Codebase Context to
jCodeMunch, Aider repo map, Repomix, and ripgrep, and checked-in
evidence plus benchmark.md reporting including known losses.
…c-integration)

Conflict resolutions:
- capabilities.md: merged OpenAI BASE_URL + Ollama provider bullets
- types.ts: use resolveDefaultApiEndpoint; drop duplicate resolveOpenAiApiEndpoint block
- embedding-config-env.test.ts: kept #146 Ollama tests with #145 OPENAI_BASE_URL stubs
- openai-embeddings.test.ts: kept #146 test title
…c-integration)

Conflict resolution: docs/benchmark.md — kept #155 front-end matrix + #156 niche benchmark sections.
…c-integration)

Conflict resolutions: README/cli/capabilities merged six-client init + 2.3.0 release copy.
Post-#156 alignment: README/CHANGELOG cite niche 50.0% top-3 and 0/6 team-convention file hits per docs/benchmark.md.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants