Skip to content

docs: separate offline quick start from model setup #10

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
rustfuture merged 1 commit into from
Oct 2, 2026
Merged

docs: separate offline quick start from model setup #10

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
17 changes: 14 additions & 3 deletions README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -15,9 +15,12 @@ Repository Intelligence searches local code and returns verbatim source lines fo

## Quick start

Requires Rust 1.85+. Offline commands use a deterministic `HashEmbedding` provider without external services. Neural embeddings and model-assisted answers use a local Ollama service (`nomic-embed-text` and `qwen2.5-coder:1.5b`) or the AGY adapter (the Google Antigravity command-line client).
You need Git and Rust 1.85+ with Cargo ([rustup](https://rustup.rs/)). The first build downloads Cargo dependencies. Offline commands use a deterministic `HashEmbedding` provider without external services. Neural embeddings and model-assisted answers use a local Ollama service (`nomic-embed-text` and `qwen2.5-coder:1.5b`) or the AGY adapter (the Google Antigravity command-line client).

```sh
git clone https://github.com/rustfuture/repository-intelligence.git
cd repository-intelligence

# Build binary
cargo build --locked

Expand All @@ -33,11 +36,19 @@ cargo run --locked -- --hybrid evaluation/corpus "index header"

# 3. Query a saved snapshot
cargo run --locked -- --load-index /tmp/ri.ri "index header"
```

The offline examples print file and line references from the bundled corpus; the snapshot is written to `/tmp/ri.ri`. No API key, running model service, or model weights are needed for those examples. Run them from the repository root. These commands use macOS/Linux paths; choose a local snapshot path on Windows.

# 4. Extractive answer: the model returns evidence IDs, the app renders the source (requires local Ollama)
### Optional model-assisted answers

Start Ollama and download `qwen2.5-coder:1.5b` before running these commands. See [the Ollama quickstart](https://docs.ollama.com/quickstart) for installation. This step needs model weights and is separate from the offline examples.

```sh
# Extractive answer: the model returns evidence IDs, the app renders the source
USE_OLLAMA=1 cargo run --locked -- --answer evaluation/corpus "What static string does reload return in api.rs?"

# 5. Machine-readable answer record (requires local Ollama)
# Machine-readable answer record
USE_OLLAMA=1 cargo run --locked -- --answer-json evaluation/corpus "What static string does reload return in api.rs?"
```

Expand Down
Loading