feat(simulation): add driver controlled backend LLM (#35186)

This commit is contained in:
James Long
2026-07-03 14:19:25 -04:00
committed by GitHub
parent 4af57079a9
commit 392cb3b080
25 changed files with 1191 additions and 21 deletions
@@ -51,6 +51,28 @@ Out of scope:
Goal: make the app safe and controlled by swapping the lowest layers, not app logic.
Implementation checklist:
- [x] Add `packages/simulation/src/backend` as the home for backend simulation layer replacements, exported from `backend/index.ts` as `simulationReplacements`; `@opencode-ai/simulation` is private/non-published and depends on logic/framework packages (`core`, `llm`, `effect`, OpenTUI), while `server` and `tui` consume it.
- [x] Wire simulation replacements through the server's `makeRoutes` via `Layer.unwrap` + dynamic `import("@opencode-ai/simulation/backend")` gated on `OPENCODE_SIMULATION`, so the simulation module is never loaded eagerly and `makeRoutes` stays synchronous.
- [x] Implement in-memory `FileSystem.FileSystem` (`simulation/filesystem.ts`) replacing the `NodeFileSystem` platform node. Backed by a flat path map; implements the operations the app uses (stat, access, chmod, realPath, read/write file, make/read directory, remove, rename, copy, copyFile, temp dirs, read-only open handles); unused operations die with a clear defect; `watch` fails as unsupported.
- [x] Root the fake filesystem at `OPENCODE_SIMULATION_ROOT` (falling back to `process.cwd()` at layer-build time). The anchor is a real, empty host directory the runner creates and cds into.
- [x] Deny host filesystem escapes loudly: content/mutation operations outside the root fail with `PermissionDenied` simulation errors. Probe operations (`stat`/`access`/`exists`) report `NotFound` outside the root so walk-up loops (project discovery, `findUp`, `globUp`) terminate naturally.
- [x] Add `SimulationFSUtil` replacement (`simulation/fs-util.ts`): wraps the real `FSUtil` layer and reroutes `readDirectoryEntries`, `glob`, and `globUp` — which bypass the injected `FileSystem` via node `fs/promises` and the `glob` package — through the simulated filesystem.
- [x] Fix `LayerNode.hoist` conflict detection to compare node implementations instead of object identity; replacement rewriting produces dependency-rewritten copies of the same node, which previously false-positived as "conflicting implementations".
- [x] Add snapshot seeding from `OPENCODE_SIMULATION_STATE`: `project/` contents of the snapshot directory are read from the host once at layer-build time and seeded into the in-memory tree joined onto the anchor root.
- [x] Verify end to end: `opencode serve` boots with `OPENCODE_SIMULATION=1` + `OPENCODE_SIMULATION_ROOT` + path/DB env seams (`OPENCODE_CONFIG_DIR`, `OPENCODE_TEST_HOME`, `OPENCODE_DB=:memory:`); `fs.list`/`fs.read` observe only seeded in-memory files; the anchor directory on the host remains empty after the run.
- [ ] Create the anchor directory + `chdir` + env seam setup automatically in CLI startup when simulation mode is enabled (currently set manually by the runner; a full run needs `OPENCODE_SIMULATION_ROOT/STATE`, `OPENCODE_CONFIG_DIR`, `OPENCODE_TEST_HOME`, `OPENCODE_DB=:memory:`, and `XDG_*_HOME` pointed into the anchor, plus Bun's `--preload=@opentui/solid/preload` when launched outside `packages/cli`).
- [ ] Assert the anchor directory is still empty at the end of the run (KV/log/flock still write through real XDG paths; they are contained in the anchor by the env seams but not yet in-memory).
- [x] Add simulated network registry (`packages/simulation/src/backend/network.ts`): replaces the `httpClient` platform node, resolves all outbound HTTP against an in-memory route table, denies unknown destinations loudly, and keeps a bounded request log (design: `simulated-network-llm.md`).
- [x] Add driver-answered LLM as an OpenAI route in the simulated network (`openai.ts` + `llm-exchange.ts`): provider requests open exchanges; the driver streams chunks back which are encoded as real OpenAI Chat SSE (schema-checked against `OpenAIChatEvent`) and consumed by the real protocol pipeline. No enqueue store — the driver is the model.
- [x] Add backend-hosted simulation control WebSocket (`control.ts`): JSON-RPC on `127.0.0.1:40950+`, started when the simulation module loads. Drivers connect directly (standalone topology — no frontend proxy): `llm.attach` (replays pending exchanges), `llm.chunk`, `llm.finish`, `llm.pending`, `network.log`; `llm.request` notifications push opened exchanges. This is also the headless-simulation interface. Drivers manage two sockets: TUI control (40900+) for UI, backend control (40950+) for LLM/network.
- [x] Answer `https://models.dev/api.json` with an empty catalog in the simulated network; providers come from seeded config (`opencode.json` in the snapshot defines an openai-compatible provider with a dummy `apiKey`, which passes the catalog availability gate and resolves onto the real openai-chat route).
- [x] Fix `buildLocationServiceMap` to apply replacements when compiling hoisted global nodes; platform-node replacements (filesystem, httpClient) were silently ignored inside hoisted globals.
- [x] Verify end to end headless (real route stack in-process + backend control WS: prompt -> `llm.request` -> driver chunks -> assistant message contains driver text; script: `packages/server/script/e2e-sim.ts`) and through the TUI (fake renderer, both sockets: type + submit via TUI WS, answer `llm.request` via backend WS, assistant reply rendered on screen; script: `packages/tui/script/sim-llm-driver.ts`).
- [ ] Add simulated process registry (shell via `just-bash`, minimal fake `git`, deny unsupported spawns).
- [ ] Trace filesystem, process, and LLM exchange activity (network requests are traced in the backend network log ring buffer; LLM exchange trace records moved out with the frontend proxy and need re-adding on the backend control server).
Scope:
- Wire simulation replacements through `AppNodeBuilder.build(...)` and `AppNodeBuilderV1.build(...)`.