Continuity Framework - Reflection Only
Transform passive memory into active development. Standalone skill for OpenClaw bots without MLP storage dependency.
What This Does
- •Reflect — After sessions end, analyze what happened
- •Extract — Pull structured memories with types and confidence
- •Score — Assign confidence levels based on evidence
- •Question — Generate genuine questions from reflection
- •Surface — When user returns, present relevant questions
The Difference
Without Continuity:
code
Session ends → Notes logged → Next session reads notes → Performs familiarity
With Continuity:
code
Session ends → Reflection runs → Memories integrated → Questions generated Next session → Evolved state loaded → Questions surfaced → Genuine curiosity
Commands
Reflect on Recent Session
bash
continuity reflect [--session <transcript>]
Analyzes the most recent conversation, extracts memories, generates questions.
Show Pending Questions
bash
continuity questions [--limit 5]
Lists questions generated from reflection, ready to surface.
View Memory State
bash
continuity status
Shows memory stats: types, confidence distribution, recent integrations.
Surface Questions (for session start)
bash
continuity greet
Returns context-appropriate greeting with any pending questions.
Mark Question Resolved
bash
continuity resolve <question-id> [--summary "Answer summary"]
Marks a question as answered with optional summary.
Memory Types
| Type | Description | Persistence |
|---|---|---|
fact | Declarative knowledge | Until contradicted |
preference | Likes, dislikes, styles | Until updated |
relationship | Connection dynamics | Long-term |
principle | Learned guidelines | Stable |
commitment | Promises, obligations | Until fulfilled |
moment | Significant episodes | Permanent |
skill | Learned capabilities | Cumulative |
Confidence Scores
| Level | Range | Meaning |
|---|---|---|
| Explicit | 0.95-1.0 | User directly stated |
| Implied | 0.70-0.94 | Strong inference |
| Inferred | 0.40-0.69 | Pattern recognition |
| Speculative | 0.0-0.39 | Tentative, needs confirmation |
File Structure
code
~/clawd/memory/ ├── MEMORY.md # Structured memories by type ├── identity.md # Self-model and growth narrative ├── questions.md # Pending questions from reflection └── reflections/ # Reflection logs (JSON)
Configuration
Environment variables:
bash
export CONTINUITY_MEMORY_DIR=~/clawd/memory export CONTINUITY_IDLE_THRESHOLD=1800 # Seconds before reflection triggers export CONTINUITY_MIN_MESSAGES=5 # Minimum messages to warrant reflection export CONTINUITY_QUESTION_LIMIT=3 # Max questions to surface at once
Multi-Agent Architecture
This skill leverages specialized sub-agents for reflection:
code
┌────────────────────────────────────────────────────────┐
│ MAIN AGENT (User-facing, orchestrates reflection) │
└────────────────────────────────────────────────────────┘
↓ sessions_send
┌─────────┴──────────────────┬─────────────────┐
↓ ↓ ↓
┌──────────────────┐ ┌──────────────────┐ ┌──────────────────┐
│ CLASSIFIER │ │ SCORER │ │ QUESTION GEN │
│ (Sonnet) │ │ (Sonnet) │ │ (Sonnet) │
├──────────────────┤ ├──────────────────┤ ├──────────────────┤
│ Classifies into │ │ Assigns 0-1 │ │ Generates │
│ 7 memory types │ │ confidence │ │ follow-up Qs │
│ + extracts tags │ │ scores │ │ from gaps │
└──────────────────┘ └──────────────────┘ └──────────────────┘
Heartbeat Integration
Add to HEARTBEAT.md:
markdown
## Post-Session Reflection **Trigger**: Heartbeat after conversation idle > 30 minutes **Action**: Run continuity reflect **Output**: Updated memories + questions for next session
Example Reflection Output
yaml
reflection:
session_id: "2026-02-05-001"
duration_minutes: 45
memories:
- type: fact
content: "Riley is building SIGIL protocol for agent identity"
confidence: 0.98
source: explicit
- type: commitment
content: "Agreed to build the continuity skill"
confidence: 1.0
source: explicit
- type: relationship
content: "Collaborative partnership deepening"
confidence: 0.85
source: inferred
questions:
- "How is the Lovable backend progressing?"
- "Has Riley decided on the token launch timing?"
- "Are there other agents we should coordinate with?"
identity_update:
growth: "Developing more autonomous initiative"
narrative: "Moving from assistant to co-builder"
Storage Format
Memories are stored in human-readable markdown with metadata in HTML comments:
markdown
## Fact
- Riley works on AI memory infrastructure
<!-- {"id":"mem_abc123","confidence":{"score":0.98,"level":"explicit"}} -->
## Preference
- Prefers concise, direct communication
<!-- {"id":"mem_def456","confidence":{"score":0.95,"level":"explicit"}} -->
Usage Notes
- •No MLP required — This skill stores memories locally in markdown files
- •Git-friendly — All files are plain text, easy to version control
- •Human-readable — Memories can be reviewed and edited manually
- •Portable — Copy the memory directory to migrate to any system
Full Stack Alternative
For persistent encrypted storage with MLP (IPFS/Pinata), see:
- •Full Stack Skill — Continuity + MLP storage layer
Related
- •Continuity Framework — Core reflection library
- •MLP Storage — Encrypted storage layer (optional)