上游文档原文（英文）

DeepSeek TUI Architecture

This document provides an overview of the DeepSeek TUI architecture for developers and contributors.

Current boundary note (v0.8.6):

crates/tui is still the live end-user runtime for the TUI, runtime API, task manager, and tool execution loop.
Other workspace crates are being split out incrementally, but they are not yet the sole runtime source of truth.
The LSP subsystem (crates/tui/src/lsp/) is fully wired into the engine's post-tool-execution path (core/engine/lsp_hooks.rs), providing inline diagnostics after every edit_file/apply_patch/write_file.
The swarm agent system was removed in v0.8.5 in favour of sub-agents (agent_spawn) and RLM (rlm_query). No model-visible swarm tool remains in the active codebase.

High-Level Overview

┌─────────────────────────────────────────────────────────────────┐
│                         User Interface                          │
│  ┌─────────────────┐  ┌─────────────────┐  ┌────────────────┐  │
│  │   TUI (ratatui) │  │  One-shot Mode  │  │  Config/CLI    │  │
│  └────────┬────────┘  └────────┬────────┘  └────────┬───────┘  │
└───────────┼─────────────────────┼────────────────────┼──────────┘
            │                     │                    │
            ▼                     ▼                    ▼
┌─────────────────────────────────────────────────────────────────┐
│                        Core Engine                              │
│  ┌─────────────────────────────────────────────────────────┐   │
│  │                    Agent Loop (core/engine.rs)           │   │
│  │  ┌─────────┐  ┌─────────────┐  ┌──────────────────────┐ │   │
│  │  │ Session │  │ Turn Mgmt   │  │ Tool Orchestration   │ │   │
│  │  └─────────┘  └─────────────┘  └──────────────────────┘ │   │
│  └─────────────────────────────────────────────────────────┘   │
└─────────────────────────────────────────────────────────────────┘
            │                     │                    │
            ▼                     ▼                    ▼
┌─────────────────────────────────────────────────────────────────┐
│                     Tool & Extension Layer                      │
│  ┌──────────┐  ┌──────────┐  ┌─────────┐  ┌────────────────┐   │
│  │  Tools   │  │  Skills  │  │  Hooks  │  │  MCP Servers   │   │
│  │ (shell,  │  │ (plugins)│  │ (pre/   │  │  (external)    │   │
│  │  file)   │  │          │  │  post)  │  │                │   │
│  └──────────┘  └──────────┘  └─────────┘  └────────────────┘   │
└─────────────────────────────────────────────────────────────────┘
            │                     │                    │
            ▼                     ▼                    ▼
┌─────────────────────────────────────────────────────────────────┐
│                  Runtime API + Task Management                  │
│  ┌─────────────────────────────┐  ┌──────────────────────────┐  │
│  │ HTTP/SSE Runtime API        │  │ Persistent Task Manager  │  │
│  │ (runtime_api.rs)            │  │ (task_manager.rs)        │  │
│  └─────────────────────────────┘  └──────────────────────────┘  │
└─────────────────────────────────────────────────────────────────┘
            │                     │
            ▼                     ▼
┌─────────────────────────────────────────────────────────────────┐
│                        LLM Layer                                │
│  ┌──────────────────────────────────────────────────────────┐  │
│  │              LLM Client Abstraction (llm_client.rs)       │  │
│  │  ┌─────────────────┐  ┌─────────────────────────────┐    │  │
│  │  │  DeepSeek Client │  │  Compatible Client (DeepSeek)│    │  │
│  │  │   (client.rs)   │  │       (client.rs)           │    │  │
│  │  └─────────────────┘  └─────────────────────────────┘    │  │
│  └──────────────────────────────────────────────────────────┘  │
└─────────────────────────────────────────────────────────────────┘

Module Organization

Entry Point

main.rs - CLI argument parsing (clap), configuration loading, entry point routing

Core Components

core/ - Main engine components
engine.rs - Engine state, operation handling, message processing
engine/turn_loop.rs - Streaming turn loop and tool execution orchestration
engine/capacity_flow.rs - Capacity guardrail checkpoints and interventions
session.rs - Session state management
turn.rs - Turn-based conversation handling
events.rs - Event system for UI updates
ops.rs - Core operations

Configuration

config.rs - Configuration loading, profiles, environment variables
settings.rs - Runtime settings management

Workspace Crates

crates/tools - Shared tool invocation primitives, including tool result/error/capability types used by the TUI runtime.
crates/agent - Model/provider registry (ModelRegistry) for resolving model IDs to provider endpoints.
crates/app-server - HTTP/SSE + JSON-RPC app server transport for headless agent workflows.
crates/config - Config loading, profiles, environment variable precedence, CLI runtime overrides.
crates/core - Agent loop, session management, turn orchestration, capacity flow guardrails.
crates/execpolicy - Approval/sandbox policy engine for tool execution decisions.
crates/hooks - Lifecycle hooks (stdout, jsonl, webhook) for pre/post tool events.
crates/mcp - MCP client + stdio server for Model Context Protocol tool servers.
crates/protocol - Request/response framing and protocol types.
crates/secrets - OS keyring integration for API key storage.
crates/state - SQLite thread/session persistence layer.
crates/tui-core - Event-driven TUI state machine scaffold.

LLM Integration

client.rs - HTTP client for DeepSeek's documented OpenAI-compatible Chat Completions API
llm_client.rs - Abstract LLM client trait with retry logic
models.rs - Data structures for API requests/responses

DeepSeek API Endpoints

DeepSeek exposes OpenAI-compatible endpoints. The CLI uses:

https://api.deepseek.com/v1/chat/completions - normal and streaming model turns
https://api.deepseek.com/v1/models - live model discovery and health checks

https://api.deepseek.com/v1 is accepted for OpenAI SDK compatibility, and https://api.deepseek.com/beta can be configured for beta-only features such as strict tool mode, chat prefix completion, and FIM completion. The public DeepSeek docs do not document a Responses API path for this workflow; the engine drives turns through Chat Completions.

Tool System

tools/ - Built-in tool implementations
mod.rs - Tool registry and common types
shell.rs - Shell command execution
file.rs - File read/write operations
todo.rs - Checklist tools plus legacy todo aliases
tasks.rs - Model-visible durable task, gate, background shell, and PR-attempt tools
github.rs - Read-only GitHub context and guarded comment/closure tools backed by gh
automation.rs - Model-visible scheduling tools over AutomationManager
plan.rs - Planning tools
subagent.rs - Sub-agent spawning (replaces the removed agent_swarm surface)
spec.rs - Tool specifications
rlm.rs - Recursive Language Model (RLM) tool — sandboxed Python REPL with llm_query() helpers

Extension Systems

mcp.rs - Model Context Protocol client for external tool servers
skills.rs - Plugin/skill loading and execution
hooks.rs - Pre/post execution hooks with conditions

User Interface

tui/ - Terminal UI components (ratatui-based)
app.rs - Application state and message handling
ui.rs - Event handling, streaming state, and rendering logic
approval.rs - Tool approval dialog
clipboard.rs - Clipboard handling
streaming.rs - Streaming text collector
ui.rs - Legacy/simple UI utilities

LSP Integration

lsp/ - Post-edit diagnostics injection (#136)
mod.rs - LspManager — lazy per-language transport pool + config
client.rs - StdioLspTransport — JSON-RPC over stdio with didOpen/didChange/publishDiagnostics
diagnostics.rs - Diagnostic types, severity, and HTML-block renderer
registry.rs - Language detection and default server map (rust-analyzer, pyright, gopls, clangd, typescript-language-server)
Wired into the engine via core/engine/lsp_hooks.rs — called after every successful edit

Security

sandbox/ - macOS sandboxing support
mod.rs - Sandbox type definitions
policy.rs - Sandbox policy configuration
seatbelt.rs - macOS Seatbelt profile generation

Utilities

utils.rs - Common utilities
logging.rs - Logging infrastructure
compaction.rs - Context compaction for long conversations
pricing.rs - Cost estimation
prompts.rs - System prompt templates
project_doc.rs - Project documentation handling
session.rs - Session serialization
runtime_api.rs - HTTP/SSE runtime API (deepseek serve --http)
runtime_threads.rs - Durable thread/turn/item store + replayable event timeline
task_manager.rs - Durable queue, worker pool, task timelines and artifacts

Data Flow

Interactive Session

User input received in TUI
Input processed by core/engine.rs
Message sent to LLM via llm_client.rs
Response streamed back, parsed in client.rs
Tool calls extracted and executed via tools/
Hooks triggered before/after tool execution
Results aggregated and sent back to LLM
Final response rendered in TUI

Crash Recovery + Offline Queue

Before sending user input, the TUI writes a checkpoint snapshot to ~/.deepseek/sessions/checkpoints/latest.json
Startup remains fresh by default; prior sessions are resumed explicitly via --resume/--continue (or Ctrl+R in TUI)
While degraded/offline, new prompts are queued in-memory and mirrored to ~/.deepseek/sessions/checkpoints/offline_queue.json
Queue edits (/queue ...) are persisted continuously so drafts and queued prompts survive restarts
Successful turn completion clears the active checkpoint and writes a durable session snapshot
Agent/Yolo turns also take pre/post-turn side-git workspace snapshots under ~/.deepseek/snapshots/<project_hash>/<worktree_hash>/.git; /restore N and revert_turn restore file state without changing conversation history or the user's .git

Tool Execution

LLM requests tool via tool_use content block
Tool registry looks up handler
Pre-execution hooks run
Approval requested if needed (non-yolo mode)
Tool executed (possibly sandboxed on macOS)
Post-execution hooks run
Result metadata is retained on runtime item records
LSP post-edit hook (v0.8.6): if the tool was edit_file/apply_patch/write_file and LSP is enabled, the engine runs run_post_edit_lsp_hook() to collect diagnostics
Diagnostics flush (v0.8.6): before the next API request, flush_pending_lsp_diagnostics() injects any collected errors as a synthetic user message
Result returned to agent loop

Background Tasks

Client enqueues task (/task add ... or POST /v1/tasks)
task_manager.rs persists task + queue entry under ~/.deepseek/tasks
Worker picks queued task (bounded pool), transitions to running
Task creates/uses a runtime thread and starts a runtime turn
runtime_threads.rs persists thread/turn/item records + monotonic event sequence
Timeline/tool summaries/artifact references are persisted incrementally
Checklist state, verifier gates, PR attempts, and guarded GitHub events are applied from tool metadata to the active task
Final state (completed|failed|canceled) is durable and queryable via TUI/API

Model-visible durable task tools are a surface over this same manager. They do not introduce a parallel work system: task_create enqueues normal tasks, checklist_* updates task-local progress, task_gate_run and completed task_shell_wait attach verification evidence, and automation runs enqueue ordinary durable tasks.

Runtime Thread/Turn Timeline

API/TUI creates or resumes a thread (/v1/threads*)
Turn starts on the thread (/v1/threads/{id}/turns)
Engine events are mapped to item lifecycle events (item.started|item.delta|item.completed)
Interrupt/steer operations apply to the active turn only
Compaction (auto/manual) is emitted as context_compaction item lifecycle
Clients replay history and resume with /v1/threads/{id}/events?since_seq=<n>

Durable Schema Gates

session_manager.rs, runtime_threads.rs, and task_manager.rs embed schema_version on persisted records.
On load, newer schema versions are rejected with explicit errors instead of silently truncating/overwriting data.
This allows safe forward migrations and prevents corruption when binaries and stored state are out of sync.

Extension Points

Adding a New Tool

Create handler in tools/
Register in tools/registry.rs
Add tool specification (name, description, input schema)

Adding an MCP Server

Configure in ~/.deepseek/mcp.json
Server auto-discovered at startup
Tools exposed to LLM automatically

Creating a Skill

Create skill directory with SKILL.md
Define skill prompt and optional scripts
Place in ~/.deepseek/skills/

Adding Hooks

Configure in ~/.deepseek/config.toml:

[[hooks]]
event = "tool_call_before"
command = "echo 'Running tool: $TOOL_NAME'"

Key Design Decisions

Streaming-first: All LLM responses stream for responsiveness
Tool safety: Non-YOLO mode requires approval for destructive operations, including side-effectful MCP tools
Extensibility: MCP, skills, and hooks allow customization without code changes
Cross-platform: Core works on Linux/macOS/Windows, sandboxing macOS-only
Minimal dependencies: Careful dependency selection for build speed
Local-first runtime API: HTTP/SSE endpoints are intended for trusted localhost access and are served by the crates/tui runtime today

Configuration Files

~/.deepseek/config.toml - Main configuration
/etc/deepseek/managed_config.toml - Optional managed defaults layer (Unix)
/etc/deepseek/requirements.toml - Optional allowed-policy constraints (Unix)
~/.deepseek/mcp.json - MCP server configuration
~/.deepseek/skills/ - User skills directory
~/.deepseek/sessions/ - Session history
~/.deepseek/sessions/checkpoints/ - Crash checkpoint + offline queue persistence
~/.deepseek/snapshots/ - Side-git pre/post-turn workspace snapshots for /restore and revert_turn
~/.deepseek/tasks/ - Background task records, queue, timelines, artifacts
~/.deepseek/audit.log - Append-only audit events for credential + approval/elevation actions

回到顶部

DeepSeek TUI 架构文档指南

这份文档真正覆盖什么

怎么用这页