Tiered Memory System v2.2.0

---

name: tiered-memory

description: "EvoClaw Tiered Memory Architecture v2.2.0 - LLM-powered three-tier memory system with automatic daily note ingestion, structured metadata extraction, URL preservation, validation, and cloud-first sync."

version: "2.2.0"

---

# Tiered Memory System v2.2.0

> *A mind that remembers everything is as useless as one that remembers nothing. The art is knowing what to keep.* 🧠

EvoClaw-compatible three-tier memory system inspired by human cognition and PageIndex tree retrieval.

What's New in v2.2.0

🔄 **Automatic Daily Note Ingestion**

What's New in v2.1.0

🎯 **Structured Metadata Extraction**

✅ **Memory Completeness Validation**

🔍 **Enhanced Search**

🛡️ **URL Preservation**

Architecture

┌─────────────────────────────────────────────────────┐
│              AGENT CONTEXT (~8-15KB)                │
│                                                     │
│  ┌──────────┐  ┌────────────────────────────────┐  │
│  │  Tree    │  │  Retrieved Memory Nodes         │  │
│  │  Index   │  │  (on-demand, 1-3KB)            │  │
│  │  (~2KB)  │  │                                │  │
│  │          │  │  Fetched per conversation      │  │
│  │  Always  │  │  based on tree reasoning       │  │
│  │  loaded  │  │                                │  │
│  └────┬─────┘  └────────────────────────────────┘  │
│       │                                             │
└───────┼─────────────────────────────────────────────┘
        │
        │ LLM-powered tree search
        │
┌───────▼─────────────────────────────────────────────┐
│              MEMORY TIERS                           │
│                                                     │
│  🔴 HOT (5KB)      🟡 WARM (50KB)     🟢 COLD (∞)  │
│                                                     │
│  Core memory       Scored facts      Full archive  │
│  - Identity        - 30-day         - Turso DB     │
│  - Owner profile   - Decaying       - Queryable    │
│  - Active context  - On-device      - 10-year      │
│  - Lessons (20 max)                                │
│                                                     │
│  Always in         Retrieved via     Retrieved via │
│  context           tree search       tree search   │
└─────────────────────────────────────────────────────┘

Design Principles

From Human Memory

From PageIndex

Cloud-First (EvoClaw)

Memory Tiers

🔴 Hot Memory (5KB max)

**Purpose:** Core identity and active context, always in agent's context window.

**Structure:**

{
  "identity": {
    "agent_name": "Agent",
    "owner_name": "User",
    "owner_preferred_name": "User",
    "relationship_start": "2026-01-15",
    "trust_level": 0.95
  },
  "owner_profile": {
    "personality": "technical, direct communication",
    "family": ["Sarah (wife)", "Luna (daughter, 3yo)"],
    "topics_loved": ["AI architecture", "blockchain", "system design"],
    "topics_avoid": ["small talk about weather"],
    "timezone": "Australia/Sydney",
    "work_hours": "9am-6pm"
  },
  "active_context": {
    "projects": [
      {
        "name": "EvoClaw",
        "description": "Self-evolving agent framework",
        "status": "Active - BSC integration for hackathon"
      }
    ],
    "events": [
      {"text": "Hackathon deadline Feb 15", "timestamp": 1707350400}
    ],
    "tasks": [
      {"text": "Deploy to BSC testnet", "status": "pending", "timestamp": 1707350400}
    ]
  },
  "critical_lessons": [
    {
      "text": "Always test on testnet before mainnet",
      "category": "blockchain",
      "importance": 0.9,
      "timestamp": 1707350400
    }
  ]
}

**Auto-pruning:**

**Generates:** `MEMORY.md` — auto-rebuilt from structured hot state

🟡 Warm Memory (50KB max, 30-day retention)

**Purpose:** Recent distilled facts with decay scoring.

**Entry format:**

{
  "id": "abc123def456",
  "text": "Decided to use zero go-ethereum deps for EvoClaw to keep binary small",
  "category": "projects/evoclaw/architecture",
  "importance": 0.8,
  "created_at": 1707350400,
  "access_count": 3,
  "score": 0.742,
  "tier": "warm"
}

**Scoring:**

score = importance × recency_decay(age) × reinforcement(access_count)

recency_decay(age) = exp(-age_days / 30)
reinforcement(access) = 1 + 0.1 × access_count

**Tier classification:**

**Eviction triggers:**

1. Age > 30 days AND score < 0.3

2. Total warm size > 50KB (evicts lowest-scored)

3. Manual consolidation

🟢 Cold Memory (Unlimited, Turso)

**Purpose:** Long-term archive, queryable but never bulk-loaded.

**Schema:**

CREATE TABLE cold_memories (
  id TEXT PRIMARY KEY,
  agent_id TEXT NOT NULL,
  text TEXT NOT NULL,
  category TEXT NOT NULL,
  importance REAL DEFAULT 0.5,
  created_at INTEGER NOT NULL,
  access_count INTEGER DEFAULT 0
);

CREATE TABLE critical_state (
  agent_id TEXT PRIMARY KEY,
  data TEXT NOT NULL,  -- {hot_state, tree_nodes, timestamp}
  updated_at INTEGER NOT NULL
);

**Retention:** 10 years (configurable)

**Cleanup:** Monthly consolidation removes frozen entries older than retention period

Tree Index

**Purpose:** Hierarchical category map for O(log n) retrieval.

**Constraints:**

**Example:**

Memory Tree Index
==================================================
📂 Root (warm:15, cold:234)
  📁 owner — Owner profile and preferences
     Memories: warm=5, cold=89
  📁 projects — Active projects
     Memories: warm=8, cold=67
    📁 projects/evoclaw — EvoClaw framework
       Memories: warm=6, cold=45
      📁 projects/evoclaw/bsc — BSC integration
         Memories: warm=3, cold=12
  📁 technical — Technical setup and config
     Memories: warm=2, cold=34
  📁 lessons — Learned lessons and rules
     Memories: warm=0, cold=44

Nodes: 7/50
Size: 1842 / 2048 bytes

**Operations:**

Distillation Engine

**Purpose:** Three-stage compression of conversations.

**Pipeline:**

Raw conversation (500B)
  ↓ Stage 1→2: Extract structured info
Distilled fact (80B)
  ↓ Stage 2→3: Generate one-line summary
Core summary (20B)

Stage 1→2: Raw → Distilled

**Input:** Raw conversation text

**Output:** Structured JSON

{
  "fact": "User decided to use raw JSON-RPC for BSC to avoid go-ethereum dependency",
  "emotion": "determined",
  "people": ["User"],
  "topics": ["blockchain", "architecture", "dependencies"],
  "actions": ["decided to use raw JSON-RPC", "avoid go-ethereum"],
  "outcome": "positive"
}

**Modes:**

**Usage:**

# Rule-based (default)
distiller.py --text "Had a productive chat about the BSC integration..." --mode rule

# LLM-powered
distiller.py --text "..." --mode llm --llm-endpoint http://localhost:8080/complete

# With core summary
distiller.py --text "..." --mode rule --core-summary

Stage 2→3: Distilled → Core Summary

**Purpose:** One-line summary for tree index

**Example:**

Distilled: {
  "fact": "User decided raw JSON-RPC for BSC, no go-ethereum",
  "outcome": "positive"
}

Core summary: "BSC integration: raw JSON-RPC (no deps)"

**Target:** <30 bytes

LLM-Powered Tree Search

**Purpose:** Semantic search through tree structure using LLM reasoning.

**How it works:**

1. **Build prompt** with tree structure + query

2. **LLM reasons** about which categories are relevant

3. **Returns** category paths with relevance scores

4. **Fetches** memories from those categories

**Example:**

Query: *"What did we decide about the hackathon deadline?"*

**Keyword search** returns:

**LLM search** reasons:

**LLM prompt template:**

You are a memory retrieval system. Given a memory tree index and a query, 
identify which categories are relevant.

Memory Tree Index:
  projects/evoclaw — EvoClaw framework (warm:6, cold:45)
  projects/evoclaw/bsc — BSC integration (warm:3, cold:12)
  ...

User Query: What did we decide about the hackathon deadline?

Output (JSON):
[
  {"path": "projects/evoclaw/bsc", "relevance": 0.95, "reason": "BSC work for hackathon"},
  {"path": "active_context/events", "relevance": 0.85, "reason": "deadline tracking"}
]

**Usage:**

# Keyword search (fast)
tree_search.py --query "BSC integration" --tree-file memory-tree.json --mode keyword

# LLM search (accurate)
tree_search.py --query "what did we decide about hackathon?" \
  --tree-file memory-tree.json --mode llm --llm-endpoint http://localhost:8080/complete

# Generate prompt for external LLM
tree_search.py --query "..." --tree-file memory-tree.json \
  --mode llm --llm-prompt-file prompt.txt

Multi-Agent Support

**Agent ID scoping** — All operations support `--agent-id` flag.

**File layout:**

memory/
  default/
    warm-memory.json
    memory-tree.json
    hot-memory-state.json
    metrics.json
  agent-2/
    warm-memory.json
    memory-tree.json
    ...
MEMORY.md              # default agent
MEMORY-agent-2.md      # agent-2

**Cold storage:** Agent-scoped queries via `agent_id` column

**Usage:**

# Store for agent-2
memory_cli.py store --text "..." --category "..." --agent-id agent-2

# Retrieve for agent-2
memory_cli.py retrieve --query "..." --agent-id agent-2

# Consolidate agent-2
memory_cli.py consolidate --mode daily --agent-id agent-2

Consolidation Modes

**Purpose:** Periodic memory maintenance and optimization.

Quick (hourly)

Daily

Monthly

Full

**Usage:**

# Quick consolidation (default)
memory_cli.py consolidate

# Daily (run via cron)
memory_cli.py consolidate --mode daily

# Monthly (run via cron)
memory_cli.py consolidate --mode monthly --db-url "$TURSO_URL" --auth-token "$TURSO_TOKEN"

**Recommended schedule:**

Critical Sync (Cloud-First)

**Purpose:** Cloud backup of hot state + tree after every conversation.

**What syncs:**

**Recovery:** If device lost, restore from cloud in <2 minutes

**Usage:**

# Manual critical sync
memory_cli.py sync-critical --db-url "$TURSO_URL" --auth-token "$TURSO_TOKEN" --agent-id default

# Automatic: Call after every important conversation
# In agent code:
#   1. Process conversation
#   2. Store distilled facts
#   3. Call sync-critical

**Retry strategy:** Exponential backoff if cloud unreachable (5s, 10s, 20s, 40s)

Metrics & Observability

**Tracked metrics:**

{
  "tree_index_size_bytes": 1842,
  "tree_node_count": 37,
  "hot_memory_size_bytes": 4200,
  "warm_memory_count": 145,
  "warm_memory_size_kb": 38.2,
  "retrieval_count": 234,
  "evictions_today": 12,
  "reinforcements_today": 67,
  "consolidation_count": 8,
  "last_consolidation": 1707350400,
  "context_tokens_saved": 47800,
  "timestamp": "2026-02-10T14:30:00"
}

**Usage:**

memory_cli.py metrics --agent-id default

**Key metrics:**

Commands Reference

Store

memory_cli.py store --text "Fact text" --category "path/to/category" [--importance 0.8] [--agent-id default]

**Importance guide:**

**Example:**

memory_cli.py store \
  --text "Decided to deploy EvoClaw on BSC testnet before mainnet" \
  --category "projects/evoclaw/deployment" \
  --importance 0.85 \
  --db-url "$TURSO_URL" --auth-token "$TURSO_TOKEN"

# Store with explicit metadata (v2.1.0+)
memory_cli.py store \
  --text "Z-Image ComfyUI model for photorealistic images" \
  --category "tools/image-generation" \
  --importance 0.8 \
  --url "https://docs.comfy.org/tutorials/image/z-image/z-image" \
  --command "huggingface-cli download Tongyi-MAI/Z-Image" \
  --path "/home/user/models/"

Validate (v2.1.0)

memory_cli.py validate [--file PATH] [--agent-id default]

**Purpose:** Check daily notes for incomplete information (missing URLs, commands, next steps).

**Example:**

# Validate today's daily notes
memory_cli.py validate

# Validate specific file
memory_cli.py validate --file memory/2026-02-13.md

**Output:**

{
  "status": "warning",
  "warnings_count": 2,
  "warnings": [
    "Tool 'Z-Image' mentioned without URL/documentation link",
    "Action 'install' mentioned without command example"
  ],
  "suggestions": [
    "Add URLs for mentioned tools/services",
    "Include command examples for setup/installation steps",
    "Document next steps after decisions"
  ]
}

Extract Metadata (v2.1.0)

memory_cli.py extract-metadata --file PATH

**Purpose:** Extract structured metadata (URLs, commands, paths) from a file.

**Example:**

memory_cli.py extract-metadata --file memory/2026-02-13.md

**Output:**

{
  "file": "memory/2026-02-13.md",
  "metadata": {
    "urls": [
      "https://docs.comfy.org/tutorials/image/z-image/z-image",
      "https://github.com/Lightricks/LTX-Video"
    ],
    "commands": [
      "huggingface-cli download Tongyi-MAI/Z-Image",
      "git clone https://github.com/Lightricks/LTX-Video.git"
    ],
    "paths": [
      "/home/peter/ai-stack/comfyui/models",
      "./configs/ltx-video-2-config.yaml"
    ]
  },
  "summary": {
    "urls_count": 2,
    "commands_count": 2,
    "paths_count": 2
  }
}

Search by URL (v2.1.0)

memory_cli.py search-url --url FRAGMENT [--limit 5] [--agent-id default]

**Purpose:** Search facts by URL fragment.

**Example:**

# Find all facts with comfy.org URLs
memory_cli.py search-url --url "comfy.org"

# Find GitHub repos
memory_cli.py search-url --url "github.com" --limit 10

**Output:**

{
  "query": "comfy.org",
  "results_count": 1,
  "results": [
    {
      "id": "abc123",
      "text": "Z-Image ComfyUI model for photorealistic images",
      "category": "tools/image-generation",
      "metadata": {
        "urls": ["https://docs.comfy.org/tutorials/image/z-image/z-image"],
        "commands": ["huggingface-cli download Tongyi-MAI/Z-Image"],
        "paths": []
      }
    }
  ]
}

Retrieve

memory_cli.py retrieve --query "search query" [--limit 5] [--llm] [--llm-endpoint URL] [--agent-id default]

**Modes:**

**Example:**

# Keyword search
memory_cli.py retrieve --query "BSC deployment decision" --limit 5

# LLM search (more accurate)
memory_cli.py retrieve \
  --query "what did we decide about blockchain integration?" \
  --llm --llm-endpoint http://localhost:8080/complete \
  --db-url "$TURSO_URL" --auth-token "$TURSO_TOKEN"

Distill

memory_cli.py distill --text "raw conversation" [--llm] [--llm-endpoint URL]

**Example:**

# Rule-based distillation
memory_cli.py distill --text "User: Let's deploy to testnet first. Agent: Good idea, safer that way."

# LLM distillation
memory_cli.py distill \
  --text "Long conversation with nuance..." \
  --llm --llm-endpoint http://localhost:8080/complete

**Output:**

{
  "distilled": {
    "fact": "Decided to deploy to testnet before mainnet",
    "emotion": "cautious",
    "people": [],
    "topics": ["deployment", "testnet", "safety"],
    "actions": ["deploy to testnet"],
    "outcome": "positive"
  },
  "mode": "rule",
  "original_size": 87,
  "distilled_size": 156
}

Hot Memory

# Update hot state
memory_cli.py hot --update KEY JSON [--agent-id default]

# Rebuild MEMORY.md
memory_cli.py hot --rebuild [--agent-id default]

# Show current hot state
memory_cli.py hot [--agent-id default]

**Keys:**

**Examples:**

# Update owner profile
memory_cli.py hot --update owner_profile '{"timezone": "Australia/Sydney", "work_hours": "9am-6pm"}'

# Add lesson
memory_cli.py hot --update lesson '{"text": "Always test on testnet first", "category": "blockchain", "importance": 0.9}'

# Add project
memory_cli.py hot --update project '{"name": "EvoClaw", "status": "Active", "description": "Self-evolving agent framework"}'

# Rebuild MEMORY.md
memory_cli.py hot --rebuild

Tree

# Show tree
memory_cli.py tree --show [--agent-id default]

# Add node
memory_cli.py tree --add "path/to/category" "Description" [--agent-id default]

# Remove node
memory_cli.py tree --remove "path/to/category" [--agent-id default]

# Prune dead nodes
memory_cli.py tree --prune [--agent-id default]

**Examples:**

# Add category
memory_cli.py tree --add "projects/evoclaw/bsc" "BSC blockchain integration"

# Remove empty category
memory_cli.py tree --remove "old/unused/path"

# Prune dead nodes (60+ days no activity)
memory_cli.py tree --prune

Cold Storage

# Initialize Turso tables
memory_cli.py cold --init --db-url URL --auth-token TOKEN

# Query cold storage
memory_cli.py cold --query "search term" [--limit 10] [--agent-id default] --db-url URL --auth-token TOKEN

**Examples:**

# Init tables (once)
memory_cli.py cold --init --db-url "https://your-db.turso.io" --auth-token "your-token"

# Query cold archive
memory_cli.py cold --query "blockchain decision" --limit 10 --db-url "$TURSO_URL" --auth-token "$TURSO_TOKEN"

Configuration

**File:** `config.json` (optional, uses defaults if not present)

{
  "agent_id": "default",
  "hot": {
    "max_bytes": 5120,
    "max_lessons": 20,
    "max_events": 10,
    "max_tasks": 10
  },
  "warm": {
    "max_kb": 50,
    "retention_days": 30,
    "eviction_threshold": 0.3
  },
  "cold": {
    "backend": "turso",
    "retention_years": 10
  },
  "scoring": {
    "half_life_days": 30,
    "reinforcement_boost": 0.1
  },
  "tree": {
    "max_nodes": 50,
    "max_depth": 4,
    "max_size_bytes": 2048
  },
  "distillation": {
    "aggression": 0.7,
    "max_distilled_bytes": 100,
    "mode": "rule"
  },
  "consolidation": {
    "warm_eviction": "hourly",
    "tree_prune": "daily",
    "tree_rebuild": "monthly"
  }
}

Integration with OpenClaw Agents

After Conversation

import subprocess
import json

def process_conversation(user_message, agent_response, category="conversations"):
    # 1. Distill conversation
    text = f"User: {user_message}\nAgent: {agent_response}"
    result = subprocess.run(
        ["python3", "skills/tiered-memory/scripts/memory_cli.py", "distill", "--text", text],
        capture_output=True, text=True
    )
    distilled = json.loads(result.stdout)
    
    # 2. Determine importance
    importance = 0.7 if "decision" in distilled["distilled"]["outcome"] else 0.5
    
    # 3. Store
    subprocess.run([
        "python3", "skills/tiered-memory/scripts/memory_cli.py", "store",
        "--text", distilled["distilled"]["fact"],
        "--category", category,
        "--importance", str(importance),
        "--db-url", os.getenv("TURSO_URL"),
        "--auth-token", os.getenv("TURSO_TOKEN")
    ])
    
    # 4. Critical sync
    subprocess.run([
        "python3", "skills/tiered-memory/scripts/memory_cli.py", "sync-critical",
        "--db-url", os.getenv("TURSO_URL"),
        "--auth-token", os.getenv("TURSO_TOKEN")
    ])

Before Responding (Retrieval)

def get_relevant_context(query):
    result = subprocess.run(
        [
            "python3", "skills/tiered-memory/scripts/memory_cli.py", "retrieve",
            "--query", query,
            "--limit", "5",
            "--llm",
            "--llm-endpoint", "http://localhost:8080/complete",
            "--db-url", os.getenv("TURSO_URL"),
            "--auth-token", os.getenv("TURSO_TOKEN")
        ],
        capture_output=True, text=True
    )
    
    memories = json.loads(result.stdout)
    return "\n".join([f"- {m['text']}" for m in memories])

Heartbeat Consolidation

import schedule

# Hourly quick consolidation
schedule.every(2).hours.do(lambda: subprocess.run([
    "python3", "skills/tiered-memory/scripts/memory_cli.py", "consolidate",
    "--mode", "quick",
    "--db-url", os.getenv("TURSO_URL"),
    "--auth-token", os.getenv("TURSO_TOKEN")
]))

# Daily tree prune
schedule.every().day.at("00:00").do(lambda: subprocess.run([
    "python3", "skills/tiered-memory/scripts/memory_cli.py", "consolidate",
    "--mode", "daily",
    "--db-url", os.getenv("TURSO_URL"),
    "--auth-token", os.getenv("TURSO_TOKEN")
]))

# Monthly full consolidation
schedule.every().month.do(lambda: subprocess.run([
    "python3", "skills/tiered-memory/scripts/memory_cli.py", "consolidate",
    "--mode", "monthly",
    "--db-url", os.getenv("TURSO_URL"),
    "--auth-token", os.getenv("TURSO_TOKEN")
]))

LLM Integration

Model Recommendations

**For Distillation & Tree Search:**

**For Tree Rebuilding:**

Cost Optimization

1. **Use cheaper models** for frequent operations (distill, search)

2. **Batch distillation** — Queue conversations, distill in batch

3. **Cache tree prompts** — Tree structure doesn't change often

4. **Skip LLM for simple** — Use rule-based for short conversations

Example LLM Endpoint

from flask import Flask, request, jsonify

app = Flask(__name__)

@app.route("/complete", methods=["POST"])
def complete():
    data = request.json
    prompt = data["prompt"]
    
    # Call your LLM (OpenAI, Anthropic, local model, etc.)
    response = llm_client.complete(prompt)
    
    return jsonify({"text": response})

if __name__ == "__main__":
    app.run(port=8080)

Performance Characteristics

**Context Size:**

**Retrieval Speed:**

**5-Year Scenario:**

Comparison with Alternatives

| System | Memory Model | Scaling | Accuracy | Cost |

|--------|-------------|---------|----------|------|

| **Flat MEMORY.md** | Linear text | ❌ Months | ⚠️ Degrades | ❌ Linear |

| **Vector RAG** | Embeddings | ✅ Years | ⚠️ Similarity≠relevance | ⚠️ Moderate |

| **EvoClaw Tiered** | Tree + tiers | ✅ Decades | ✅ Reasoning-based | ✅ Fixed |

**Why tree > vectors:**

Troubleshooting

Tree size exceeding limit

# Prune dead nodes
memory_cli.py tree --prune

# Check which nodes are largest
memory_cli.py tree --show | grep "Memories:"

# Manually remove unused categories
memory_cli.py tree --remove "unused/category"

Warm memory filling up

# Run consolidation
memory_cli.py consolidate --mode daily --db-url "$TURSO_URL" --auth-token "$TURSO_TOKEN"

# Check stats
memory_cli.py metrics

# Lower eviction threshold (keeps less in warm)
# Edit config.json: "eviction_threshold": 0.4

Hot memory exceeding 5KB

# Hot auto-prunes, but check structure
memory_cli.py hot

# Remove old projects/tasks manually
memory_cli.py hot --update project '{"name": "OldProject", "status": "Completed"}'

# Rebuild to force pruning
memory_cli.py hot --rebuild

LLM search failing

# Fallback to keyword search (automatic)
memory_cli.py retrieve --query "..." --limit 5

# Test LLM endpoint
curl -X POST http://localhost:8080/complete -d '{"prompt": "test"}'

# Generate prompt for external testing
tree_search.py --query "..." --tree-file memory/memory-tree.json --mode llm --llm-prompt-file test.txt

Migration from v1.x

**Backward compatible:** Existing `warm-memory.json` and `memory-tree.json` files work as-is.

**New files:**

**Steps:**

1. Update skill: `clawhub update tiered-memory`

2. Run consolidation to rebuild hot state: `memory_cli.py consolidate`

3. Initialize cold storage (optional): `memory_cli.py cold --init --db-url ... --auth-token ...`

4. Configure agent to use new commands (see Integration section)

Migration from v2.0 to v2.1

**Fully backward compatible:** Existing memory files work without changes.

**What's new:**

**Testing the upgrade:**

# Update skill
clawhub update tiered-memory

# Test metadata extraction
memory_cli.py extract-metadata --file memory/2026-02-13.md

# Validate your recent notes
memory_cli.py validate

# Search by URL
memory_cli.py search-url --url "github.com"

References

---

*v2.1.0 — A mind that remembers everything is as useless as one that remembers nothing. The art is knowing what to keep. Now with structured metadata to remember HOW, not just WHAT.* 🧠🌲🔗