46424cfbc4
- Add state.rs with AgentState/Step/StepStatus/AgentPhase as single source of truth - Extract prompts to markdown files loaded via include_str! - Replace plan_steps table with execution_log + agent_state_snapshots - Implement user feedback processing with docker-build-cache plan diff: load snapshot → LLM revise_plan → diff (title, description) → invalidate from first mismatch → resume - run_agent_loop accepts optional initial_state for mid-execution resume - Broadcast plan step status (done/running/pending) to frontend on step transitions - Rewrite frontend types/components to match new API (ExecutionLogEntry, PlanStepInfo with status) Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
87 lines
2.4 KiB
Markdown
87 lines
2.4 KiB
Markdown
# Agent Context 构建
|
||
|
||
AgentState 如何变成 LLM API call 的 messages 数组。定义见 `src/state.rs`。
|
||
|
||
## AgentState
|
||
|
||
```rust
|
||
struct AgentState {
|
||
phase: AgentPhase, // Planning | Executing { step } | Completed
|
||
steps: Vec<Step>, // 执行计划，每个 step 有 status + optional summary
|
||
current_step_chat_history: Vec<ChatMessage>, // 当前步骤内的多轮对话，step 切换时 clear
|
||
scratchpad: String, // LLM 的跨步骤工作区
|
||
}
|
||
```
|
||
|
||
整个结构体 `Serialize/Deserialize`，可 JSON 直接存 DB。
|
||
|
||
## Step 生命周期
|
||
|
||
```
|
||
Pending → Running → Done(summary) 或 Failed
|
||
```
|
||
|
||
summary 在 step 完成时由 LLM 填入，作为后续步骤的压缩上下文。
|
||
|
||
## Messages 组装
|
||
|
||
### Planning 阶段
|
||
|
||
```
|
||
[ system(planning_prompt), user(requirement), ...current_step_chat_history ]
|
||
```
|
||
|
||
### Executing 阶段
|
||
|
||
```
|
||
[ system(execution_prompt), user(step_context), ...current_step_chat_history ]
|
||
```
|
||
|
||
`step_context` 由 `build_step_context()` 拼接：
|
||
|
||
```
|
||
## 需求
|
||
{requirement}
|
||
|
||
## 计划概览
|
||
1. 分析代码结构 done
|
||
2. 实现核心逻辑 >> current
|
||
3. 写测试 FAILED
|
||
4. 集成测试
|
||
|
||
## 当前步骤（步骤 2）
|
||
标题：实现核心逻辑
|
||
描述：...
|
||
|
||
## 已完成步骤摘要 ← 从 steps 中 filter Done，取 summary
|
||
- 步骤 1: ...
|
||
|
||
## Scratchpad ← LLM 自己维护的跨步骤笔记
|
||
...
|
||
```
|
||
|
||
## 持久化
|
||
|
||
### DB 表
|
||
|
||
- **agent_state_snapshots** — step 切换时 insert 一行 AgentState JSON 快照（追加，不覆盖，保留历史）。恢复时取最新一行。
|
||
- **execution_log** — tool call 的输入输出记录（不可变历史），前端展示 + report 生成用。
|
||
|
||
Plan 步骤只从 AgentState JSON 读，不再单独写表。
|
||
|
||
### Executing 阶段 text response
|
||
|
||
LLM 返回纯文本时不隐含"workflow 结束"，写 execution_log 显示给用户。只有显式调 `advance_step` 才推进步骤。
|
||
|
||
## 待做：Context 压缩
|
||
|
||
当前无长度限制，`current_step_chat_history` 在单步 tool call 轮次过多时会无限增长。
|
||
|
||
压缩按优先级从低到高砍：
|
||
|
||
| 优先级 | 内容 | 处理 |
|
||
|---|---|---|
|
||
| 高 | system prompt / requirement / plan 概览 | 不压缩 |
|
||
| 中 | 当前步骤最近 N 轮 tool call | 完整保留 |
|
||
| 低 | 当前步骤早期 tool call | 替换为摘要 |
|