Set up Cursor Cloud dev environment - #1
Merged
Conversation
Co-authored-by: Moses Man <mosesman831@users.noreply.github.com>
There was a problem hiding this comment.
Pull request overview
This PR adds Cursor Cloud-specific development environment guidance and a local config override to let Dendrite run against a local OpenAI-compatible LLM endpoint (intended to work without API keys), without changing application source code.
Changes:
- Added
dendrite.config.local.yamlfor running Dendrite against a local OpenAI-compatible LLM endpoint (e.g., Ollama) withapiKeyEnv: NONE. - Extended
AGENTS.mdwith Cursor Cloud-specific setup/verification caveats (build-before-run, native module build, keyless checks, and local LLM guidance).
Reviewed changes
Copilot reviewed 2 out of 2 changed files in this pull request and generated 3 comments.
| File | Description |
|---|---|
| dendrite.config.local.yaml | Adds a local/offline-friendly config override for Cursor Cloud agents. |
| AGENTS.md | Documents Cursor Cloud-specific environment setup and verification guidance. |
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
Comment on lines
+12
to
+16
| stt: | ||
| provider: openai-audio | ||
| baseURL: http://127.0.0.1:11434/v1 | ||
| model: whisper-1 | ||
| apiKeyEnv: NONE |
|
|
||
| Startup runs `npm install` automatically (the update script). Everything below is durable, non-obvious context — see `README.md` / `package.json` scripts for the standard commands. | ||
|
|
||
| - **Build before running anything.** The `dendrite` bin points at `dist/cli.js`, so `npm run build` (tsc) must be run before `doctor`/`ingest`/`serve`/`mcp`/`npm test`. There is no ts-runtime; rebuild after any `src/` change. Build is intentionally kept out of the update script. |
| - **LLM is required for real work.** `ingest`, `serve`, `mcp` and the full `npm test` (31 checks) call an OpenAI-compatible chat endpoint. No `OPENAI_API_KEY`/`NVIDIA_API_KEY` is set by default. Two ways to get a working LLM: | ||
| 1. Add `OPENAI_API_KEY` and/or `NVIDIA_API_KEY` as secrets (default `dendrite.config.yaml` uses NVIDIA primary + OpenAI fallback). | ||
| 2. **Offline (no keys):** run [Ollama] serving a small model and point Dendrite at it. Use `dendrite.config.local.yaml` (already in repo) via `-c dendrite.config.local.yaml`; it targets `http://127.0.0.1:11434/v1`, model `llama3.2:3b`, `apiKeyEnv: NONE`. | ||
| - **Ollama gotcha:** the newest Ollama (0.31.x) **segfaults during model warmup on this CPU**. Install a stable older build instead: `curl -fsSL https://ollama.com/install.sh | OLLAMA_VERSION=0.6.8 sh`, then `ollama serve` (systemd isn't running, so start it manually, e.g. in tmux) and `ollama pull llama3.2:3b`. Ollama is deliberately NOT in the update script (too heavy). |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
What this does
Sets up and verifies the Dendrite development environment for Cursor Cloud agents. No application source was modified — only environment docs and a local LLM config helper were added.
Changes
AGENTS.md: added a## Cursor Cloud specific instructionssection documenting durable, non-obvious caveats (build-before-run since the bin runs fromdist/,better-sqlite3native build, no-key checks, LLM requirement, the Ollama-version segfault gotcha, index-db location).dendrite.config.local.yaml: a config override that points Dendrite at a local OpenAI-compatible LLM (http://127.0.0.1:11434/v1,llama3.2:3b,apiKeyEnv: NONE) so the pipeline can run fully offline without API keys.Update script (startup)
npm install(minimal; build/test/service startup intentionally excluded).Verification
npm install— deps +better-sqlite3native buildnpm run buildnpm run test:ci— smoke suite (no keys)node dist/cli.js doctor -c dendrite.config.local.yaml— LLM reachabledendrite ingest— classified + wrote structured Markdown notes into the vault (learnings,memories), with append-to-existing and cross-linkingnpm test— 30/31 with the small local model; the singleFTS searchfailure is a data/model-quality artifact (weak local model routing), not an env issueHello-world log:
hello_world_ingest.log
Full suite log:
full_test_local_llm.log
To show artifacts inline, enable in settings.