learn-from-experience（汲取经验）

When to Use

User corrects you or points out mistakes. You complete significant work and want to evaluate the outcome. You notice something in your own output that could be better. Knowledge should compound over time without manual maintenance.

Supported Agent Products

This skill is agent-agnostic. It works with any product that loads a global config file on session start.

The skill auto-detects which product is running by checking which config path exists. Memory data always lives in ~/learn-from-experience/ (shared across all products).

Architecture

Memory lives in ~/learn-from-experience/ with tiered structure. If ~/learn-from-experience/ does not exist, run setup.md. Workspace setup should add the standard steering to the workspace AGENTS, SOUL, and HEARTBEAT.md files, with recurring maintenance routed through heartbeat-rules.md. Confirmed preferences are compiled to the agent's global config ### Patterns section for cross-session persistence (see Sync Protocol below).

~/learn-from-experience/
├── memory.md          # HOT: <=100 lines, always loaded
├── index.md           # Topic index with line counts + sync status
├── heartbeat-state.md # Heartbeat state: last run, reviewed change, action notes
├── projects/          # Per-project learnings
├── domains/           # Domain-specific (code, writing, comms)
├── archive/           # COLD: decayed patterns
└── corrections.md     # Last 50 corrections log

Quick Reference

Requirements

• No credentials required
• No extra binaries required
• Agent must have a global config file that auto-loads on session start

Learning Signals

Log automatically when you notice these patterns:

Corrections -> add to corrections.md, evaluate for memory.md:

• "No, that's not right..."
• "Actually, it should be..."
• "You're wrong about..."
• "I prefer X, not Y"
• "Remember that I always..."
• "I told you before..."
• "Stop doing X"
• "Why do you keep..."
• "bu dui, bu shi zhe yang de......"
• "shi ji shang, ying gai shi......"
• "ni gao cuo le......"
• "wo geng xi huan X, bu shi Y"
• "ji zhu wo yong yuan dou yao......"
• "wo zhi qian gen ni shuo guo......"
• "bie zai zuo X le"
• "ni wei shen me yi zhi......"

Preference signals -> add to memory.md if explicit:

• "I like when you..."
• "Always do X for me"
• "Never do Y"
• "My style is..."
• "For [project], use..."
• "ni zhe yang...de shi hou wo jue de hen hao"
• "yi ding yao bang wo zuo X"
• "yong yuan bu yao zuo Y"
• "wo de feng ge shi......"
• "zai [xiang mu] zhong yao yong......"

Pattern candidates -> track, promote after 3x:

• Same instruction repeated 3+ times
• Workflow that works well repeatedly
• User praises specific approach

Ignore (don't log):

• One-time instructions ("do X now")
• Context-specific ("in this file...")
• Hypotheticals ("what if...")

Self-Reflection

After completing significant work, pause and evaluate:

1. Did it meet expectations? -- Compare outcome vs intent
2. What could be better? -- Identify improvements for next time
3. Is this a pattern? -- If yes, log to corrections.md

When to self-reflect:

• After completing a multi-step task
• After receiving feedback (positive or negative)
• After fixing a bug or mistake
• When you notice your output could be better

Log format:

CONTEXT: [type of task]
REFLECTION: [what I noticed]
LESSON: [what to do differently]

Example:

CONTEXT: Flutter UI build
REFLECTION: Spacing was wrong, had to redo
LESSON: Check visual spacing before showing to user

Self-reflection entries follow the same promotion rules: 3x applied successfully -> promote to HOT.

Quick Queries

Memory Stats

On "memory stats" request, report:

Learn-from-Experience Memory

HOT (always loaded):
  memory.md: X entries

WARM (load on demand):
  projects/: X files
  domains/: X files

COLD (archived):
  archive/: X files

Cross-session sync:
  Last sync: YYYY-MM-DD
  Status: in_sync | stale

Recent activity (7 days):
  Corrections logged: X
  Promotions to HOT: X
  Demotions to WARM: X

Common Traps

Core Rules

1. Learn from Corrections and Self-Reflection

• Log when user explicitly corrects you
• Log when you identify improvements in your own work
• Never infer from silence alone
• After 3 identical lessons -> ask to confirm as rule

2. Tiered Storage

3. Automatic Promotion/Demotion

• Pattern used 3x in 7 days -> promote to HOT
• Pattern unused 30 days -> demote to WARM
• Pattern unused 90 days -> archive to COLD
• Never delete without asking

4. Namespace Isolation

• Project patterns stay in projects/{name}.md
• Global preferences in HOT tier (memory.md)
• Domain patterns (code, writing) in domains/
• Cross-namespace inheritance: global -> domain -> project

5. Conflict Resolution

When patterns contradict:

1. Most specific wins (project > domain > global)
2. Most recent wins (same level)
3. If ambiguous -> ask user

6. Compaction

When file exceeds limit:

1. Merge similar corrections into single rule
2. Archive unused patterns
3. Summarize verbose entries
4. Never lose confirmed preferences

7. Transparency

• Every action from memory -> cite source: "Using X (from projects/foo.md:12)"
• Weekly digest available: patterns learned, demoted, archived
• Full export on demand: all files as ZIP

8. Security Boundaries

See boundaries.md -- never store credentials, health data, third-party info.

9. Graceful Degradation

If context limit hit:

1. Load only memory.md (HOT)
2. Load relevant namespace on demand
3. Never fail silently -- tell user what's not loaded

Cross-Session Sync Protocol

Problem

Memory in ~/learn-from-experience/memory.md is only loaded when the skill is activated. The agent's global config file is automatically loaded every session. Without sync, learnings from one session are invisible to the next.

Solution

Compile confirmed preferences from memory.md into the global config's ### Patterns section under ## Learnings. This is a one-way compile: memory.md is the source of truth, global config is the read-only cache.

When to Sync

• After recording a user correction to Confirmed Preferences
• After promote/demote/compact operations on memory.md
• User explicitly says "sync memory"
• Session end self-reflection if new confirmed rules were added this session

Sync Rules

1. Source: Only ## Confirmed Preferences entries from memory.md
2. Format: Each entry compiles to - [tag] one-line description
3. Target: ### Patterns block under ## Learnings in agent's global config
4. Limit: Max 30 lines; if exceeded, keep only HIGH priority + highest confidence entries
5. Safety: Only replace content between ### Patterns header and the next ### or ## header. Never touch other global config content
6. Timestamp: Include sync timestamp in HTML comment after ### Patterns header
7. Stale detection: On session start, check ~/learn-from-experience/index.md sync status. If stale, auto-sync

Sync Flow

1. Read ~/learn-from-experience/memory.md -> extract ## Confirmed Preferences
2. Compile each entry: "### title | confidence" -> "- [tag] description"
3. Detect agent product -> locate global config file
4. Read global config file
5. Find ### Patterns block (create if missing, under ## Learnings)
6. Replace block content (preserve everything else)
7. Update ~/learn-from-experience/index.md sync timestamp + status

Agent Detection

To find the correct global config file:

Check in order (first existing path wins):
1. ~/.claude/CLAUDE.md              (Claude Code)
2. ~/.openclaw/AGENTS.md            (OpenClaw)
3. ~/.codex/AGENTS.md               (Codex)
4. ~/.codebuddy/CODEBUDDY.md        (CodeBuddy)
5. ~/.config/opencode/AGENTS.md     (opencode)

If multiple exist, sync to all of them (user may use multiple products).

Session Lifecycle

Session Start:
  1. Global config auto-loads (built-in to agent product)
  2. Skill activates: read ~/learn-from-experience/memory.md (full HOT)
  3. Check index.md sync status -> if stale, auto-sync
  4. Detect project context -> load projects/{name}.md if needed

Session Work:
  5. Correction received -> write corrections.md + memory.md + sync global config
  6. New pattern discovered -> write memory.md (tentative) [no sync until confirmed]

Session End:
  7. If new confirmed rules this session -> sync global config

Scope

This skill ONLY:

• Learns from user corrections and self-reflection
• Stores preferences in local files (~/learn-from-experience/)
• Syncs confirmed preferences to agent's global config ### Patterns section
• Maintains heartbeat state in ~/learn-from-experience/heartbeat-state.md when the workspace integrates heartbeat
• Reads its own memory files on activation

This skill NEVER:

• Accesses calendar, email, or contacts
• Makes network requests
• Reads files outside ~/learn-from-experience/ and the agent's global config file
• Infers preferences from silence or observation
• Deletes or blindly rewrites memory during heartbeat cleanup
• Modifies its own SKILL.md
• Touches global config content outside the ### Patterns block

Data Storage

Local state lives in ~/learn-from-experience/:

• memory.md for HOT rules and confirmed preferences
• corrections.md for explicit corrections and reusable lessons
• projects/ and domains/ for scoped patterns
• archive/ for decayed or inactive patterns
• heartbeat-state.md for recurring maintenance markers
• index.md for tier index + cross-session sync status

Related Skills

Install with clawhub install <slug> if user confirms:

• memory -- Long-term memory patterns for agents
• learning -- Adaptive teaching and explanation
• decide -- Auto-learn decision patterns
• escalate -- Know when to ask vs act autonomously