Skip to content

feat: add opencode.json + fix stale ollama-arc refs in AGENTS.md #16

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
growlf merged 1 commit into from
May 11, 2026
+10 −4
Merged

feat: add opencode.json + fix stale ollama-arc refs in AGENTS.md #16

Show file tree
Hide file tree
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
8 changes: 4 additions & 4 deletions AGENTS.md
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,6 +1,6 @@
# AGENTS.md — ai-stack

This is a Docker Compose-based AI stack for **Intel Arc iGPU on Linux**, managed via systemd. The stack provides local LLM inference (ollama-arc), cloud API routing (LiteLLM), unified routing/load balancing (Olla), and Obsidian vault RAG (retriever). The primary AI interface is **OpenCode** (CLI + Obsidian sidebar plugin).
This is a Docker Compose-based AI stack managed via systemd. The stack provides local LLM inference (ollama), cloud API routing (LiteLLM), unified routing/load balancing (Olla), and Obsidian vault RAG (retriever). The primary AI interface is **OpenCode** (CLI + Obsidian sidebar plugin).

## Developer commands

Expand Down Expand Up @@ -58,7 +58,7 @@ All traffic flows through **Olla** (port 40114) as the unified LLM router:
OpenCode (CLI + Obsidian plugin)
├── tool: retriever :42000 → sqlite-vec + FTS5 hybrid search over vault
├── provider: Olla :40115 → Smart Router (auto-selects local model)
│ → ollama-arc :11434 (Intel Arc iGPU)
│ → ollama :11434 (local LLM)
│ → OLLAMA_REMOTE_* nodes (LAN, optional)
└── provider: LiteLLM :4000 → Claude (Anthropic), Gemini (Google)
```
Expand All @@ -67,7 +67,7 @@ OpenCode (CLI + Obsidian plugin)

| Directory/File | Purpose |
|---|---|
| `docker-compose.yml` | Core stack: ollama-arc, litellm, olla, router, retriever |
| `docker-compose.yml` | Core stack: ollama, litellm, olla, router, retriever |
| `install.sh` | Preflight → create volumes → install systemd → start stack → pull models (prompts for Bitwarden setup) |
| `retriever/` | Obsidian vault RAG: FastAPI + sqlite-vec + watchdog. Hybrid search via FTS5 + vector embeddings. |
| `scripts/generate-olla-config.sh` | Reads `OLLAMA_REMOTE_*` from `.env` → writes `proxy/olla.yaml` |
Expand All @@ -94,7 +94,7 @@ The retriever service replaces Khoj + PostgreSQL with a lightweight, API-only se
- **Vector store**: sqlite-vec (embedded SQLite extension, file-based, no separate DB)
- **Keyword search**: SQLite FTS5 (BM25 scoring)
- **Hybrid search**: Reciprocal Rank Fusion (RRF) combining vector + keyword results
- **Embeddings**: `nomic-embed-text` via Olla → ollama-arc
- **Embeddings**: `nomic-embed-text` via Olla → ollama
- **Indexing**: Full scan on startup, then watchdog (inotify) for live changes
- **API**: `POST /search`, `POST /reindex`, `GET /health`

Expand Down
6 changes: 6 additions & 0 deletions opencode.json
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,6 @@
{
"$schema": "https://opencode.ai/config.json",
"instructions": [
"AGENTS.md"
]
}