介绍
# Causal Inference
一个轻量级的因果层，用于预测行动结果，不是通过模式匹配相关性，而是通过建模干预和反事实。
## 核心不变量
**每个行动都必须能表示为对因果模型的显式干预，并包含预测效果 + 不确定性 + 可证伪的审计轨迹。**
计划必须是*因果有效的*，而不仅仅是看似合理的。
## 何时触发
**在 ANY 高层级行动上触发此技能**，包括但不限于：
| 领域 | 要记录的行动 | |--------|---------------| | **沟通** | 发送邮件，发送消息，回复，跟进，通知，提及 | | **日历** | 创建/移动/取消会议，设置提醒，RSVP | | **任务** | 创建/完成/推迟任务，设置优先级，指派 | | **文件** | 创建/编辑/分享文档，提交代码，部署 | | **社交** | 发布，反应，评论，分享，私信 | | **购买** | 订单，订阅，取消，退款 | | **系统** | 配置变更，权限授予，集成设置 |
在以下情况也触发： - **审查结果** — "那封邮件得到回复了吗？" → 记录结果，更新估算 - **调试失败** — "为什么这没有起作用？" → 追踪因果图 - **回填历史** — "分析我过去的邮件/日历" → 解析日志，重建行动 - **规划** — "我应该现在还是稍后发送？" → 查询因果模型
## 回填：从历史数据引导
不要从零开始。解析现有日志以重建过去的行动 + 结果。
### 邮件回填
```bash # Extract sent emails with reply status gog gmail list --sent --after 2024-01-01 --format json > /tmp/sent_emails.json
# For each sent email, check if reply exists python3 scripts/backfill_email.py /tmp/sent_emails.json ```
### 日历回填
```bash # Extract past events with attendance gog calendar list --after 2024-01-01 --format json > /tmp/events.json
# Reconstruct: did meeting happen? was it moved? attendee count? python3 scripts/backfill_calendar.py /tmp/events.json ```
### 消息回填 (WhatsApp/Discord/Slack)
```bash # Parse message history for send/reply patterns wacli search --after 2024-01-01 --from me --format json > /tmp/wa_sent.json python3 scripts/backfill_messages.py /tmp/wa_sent.json ```
### 通用回填模式
```python # For any historical data source: for record in historical_data: action_event = { "action": infer_action_type(record), "context": extract_context(record), "time": record["timestamp"], "pre_state": reconstruct_pre_state(record), "post_state": extract_post_state(record), "outcome": determine_outcome(record), "backfilled": True # Mark as reconstructed } append_to_log(action_event) ```
## 架构
### A. 行动日志 (必需)
每个执行的行动都会发出一个结构化事件：
```json { "action": "send_followup", "domain": "email", "context": {"recipient_type": "warm_lead", "prior_touches": 2}, "time": "2025-01-26T10:00:00Z", "pre_state": {"days_since_last_contact": 7}, "post_state": {"reply_received": true, "reply_delay_hours": 4}, "outcome": "positive_reply", "outcome_observed_at": "2025-01-26T14:00:00Z", "backfilled": false } ```
存储在 `memory/causal/action_log.jsonl`。
### B. 因果图 (按领域)
每个领域从 10-30 个可观察变量开始。
**邮件领域：** ``` send_time → reply_prob subject_style → open_rate recipient_type → reply_prob followup_count → reply_prob (diminishing) time_since_last → reply_prob ```
**日历领域：** ``` meeting_time → attendance_rate attendee_count → slip_risk conflict_degree → reschedule_prob buffer_time → focus_quality ```
**消息领域：** ``` response_delay → conversation_continuation message_length → response_length time_of_day → response_prob platform → response_delay ```
**任务领域：** ``` due_date_proximity → completion_prob priority_level → completion_speed task_size → deferral_risk context_switches → error_rate ```
将图定义存储在 `memory/causal/graphs/`。
### C. 估算
对于每个"旋钮"（干预变量），估算处理效应：
```python # Pseudo: effect of morning vs evening sends effect = mean(reply_prob | send_time=morning) - mean(reply_prob | send_time=evening) uncertainty = std_error(effect) ```
首先使用简单的回归或倾向匹配。当图是显式的并且需要识别时，进阶到 do-calculus。
### D. 决策策略
在执行行动之前：
1. 识别干预变量 2. 查询因果模型以获取预期的结果分布 3. 计算预期效用 + 不确定性边界 4. 如果不确定性 > 阈值 或 预期危害 > 阈值 → 拒绝或升级给用户 5. 记录预测以供后续验证
## 工作流
### 在每次行动时
``` BEFORE executing: 1. Log pre_state 2. If enough historical data: query model for expected outcome 3. If high uncertainty or risk: confirm with user
AFTER executing: 1. Log action + context + time 2. Set reminder to check outcome (if not immediate)
WHEN outcome observed: 1. Update action log with post_state + outcome 2. Re-estimate treatment effects if enough new data ```
### 规划行动时
``` 1. User request → identify candidate actions 2. For each action: a. Map to intervention(s) on causal graph b. Predict P(outcome | do(action)) c. Estimate uncertainty d. Compute expected utility 3. Rank by expected utility, filter by safety 4. Execute best action, log prediction 5. Observe outcome, update model ```
### 调试失败时
``` 1. Identify failed outcome 2. Trace back through causal graph 3. For each upstream node: a. Was the value as expected? b. Did the causal link hold? 4. Identify broken link(s) 5. Compute minimal intervention set that would have prevented failure 6. Log counterfactual for learning ```
## 快速开始：今天就引导
```bash # 1. Create the infrastructure mkdir -p memory/causal/graphs memory/causal/estimates
# 2. Initialize config cat > memory/causal/config.yaml << 'EOF' domains: - email - calendar - messaging - tasks
thresholds: max_uncertainty: 0.3 min_expected_utility: 0.1
protected_actions: - delete_email - cancel_meeting - send_to_new_contact - financial_transaction EOF
# 3. Backfill one domain (start with email) python3 scripts/backfill_email.py
# 4. Estimate initial effects python3 scripts/estimate_effect.py --treatment send_time --outcome reply_received --values morning,evening ```
## 安全约束
定义需要明确用户批准的"受保护变量"：
```yaml protected: - delete_email - cancel_meeting - send_to_new_contact - financial_transaction
thresholds: max_uncertainty: 0.3 # don't act if P(outcome) uncertainty > 30% min_expected_utility: 0.1 # don't act if expected gain < 10% ```
## 文件
- `memory/causal/action_log.jsonl` — 所有带有结果的已记录行动 - `memory/causal/graphs/` — 特定领域的因果图定义 - `memory/causal/estimates/` — 学习到的处理效应 - `memory/causal/config.yaml` — 安全阈值和受保护变量
## 参考
- 参见 `references/do-calculus.md` 了解形式化的干预语义 - 参见 `references/estimation.md` 了解处理效应估算方法