Dependency Flow

Dependencies flow downward only. A crate may depend on crates below it in the stack but never above.

rho (binary) ──────────────────────────────────────────
    │
    ├── rho-ext ──── rho-core ──── rho-ai ──────────────
    │
    ├── rho-tools ── rho-highlight ── rho-core ─────────
    │           └──── rho-memory ── rho-core ────────────
    │
    └── rho-core ──── rho-ai ──────────────────────────

rho-test-helpers ── rho-core, rho-ai ────────────────
xtask ──────────────────────────────────────────────

Crate	Depends on	Notes
`rho`	`rho-core`, `rho-tools`, `rho-ext`, `rho-ai`	Headless JSON-RPC 2.0 agent
`rho-ext`	`rho-core`, `deno_core`, `deno_ast`	TypeScript extension runtime
`rho-tools`	`rho-core`, `rho-highlight`, `rho-memory`	Tool implementations; highlight for node-splitting, memory for `MemoryTool`
`rho-memory`	`rho-core`	Persistent knowledge base (SQLite/FTS5)
`rho-highlight`	`rho-core`	Tree-sitter grammar; uses core diagnostic types
`rho-core`	`rho-ai`	Kernel — the foundation everything else builds on
`rho-ai`	none (external only)	Unified LLM provider abstraction
`rho-test-helpers`	`rho-core`, `rho-ai`	Dev-only — provides mocks and fixtures for testing
`xtask`	none (cargo integration)	Dev-only — task runner, no rho crate dependencies

Key external dependencies

Dependency	Used by	Purpose
`reqwest`	`rho-ai`	HTTP client for model API
`tokio`	`rho-core`, `rho-tools`, `rho-ext`	Async runtime
`serde` / `serde_json`	everywhere	Serialization
`toml`	`rho-core`	Configuration parsing
`tree-sitter` + grammars	`rho-highlight`	Syntax analysis
`ignore`	`rho-tools`	`.gitignore`-aware directory listing
`clap`	`rho`	CLI argument parsing
`tracing`	`rho-core`, `rho-ext`	Structured logging
`deno_core`	`rho-ext`	V8 JavaScript runtime
`deno_ast`	`rho-ext`	TypeScript transpilation
`thiserror`	`rho-core`, `rho-ext`	Error type derivation
`futures`	`rho-ai`	Stream traits for SSE
`sqlx`	`rho-memory`	SQLite + FTS5 for the knowledge base

What this buys you

The layered structure means:

rho-core is UI-agnostic — no shell, no filesystem, no terminal. Any frontend (REPL, TUI, editor integration) consumes it through the JSON-RPC 2.0 protocol.
Tools are pluggable — ToolRegistry stores Box<dyn Tool>, so adding a tool is registry.register(Box::new(MyTool)).
Testing is isolated — rho-test-helpers mocks the core traits (MockChatClient, MockShellExecutor, ApprovalGate) so tool and agent loop tests run without a model server.