Continuity Protocol

A protocol for maintaining project coherence across time gaps

The Problem

When you return to a project after days, weeks, or months away:

• "Where was I?"
• "What was I thinking?"
• "What happened while I was away?"
• "Am I still aligned with my goals?"

Existing tools don't solve this. Notes get stale. Memory fades. Context is lost.

The Solution

The Continuity Protocol defines 5 primitives that answer these questions through artifact-based reconstruction:

1. State → Where am I?
2. Sessions → What did I do?
3. Archaeology → What happened when I was away?
4. Drift Detection → Am I still aligned?
5. Specification Locks → What must not change?

Not another productivity app. A protocol that works with tools you already use.

Quick Example

# Initialize project
python continuity.py setup
python continuity.py init "My Project" --goal "Build feature X"

# Work session
python continuity.py session start
# ... work on your project ...
python continuity.py session end --summary "Completed auth flow"

# Return after 2 weeks away
python continuity.py resume

Output:

🔄 RESUMING PROJECT
📦 Project: My Project
🎯 Goal: Build feature X
📊 Total sessions: 3

⚠️  DRIFT ALERTS:
   - 14 day gap since last session - run archaeology

🔍 Recent activity:
   [Shows timeline of last 3 sessions, gaps detected]

✅ Ready to continue

How It Works

Protocol, Not Software

Like Git, Markdown, or RSS, the Continuity Protocol is a specification, not an app.

This means:

• Works on multiple platforms (Notion, GitHub, Obsidian, etc.)
• Multiple implementations can coexist
• You own your data and the pattern
• Protocol outlives any single tool

Reference Implementation

This repository contains:

• PROTOCOL.md - Technical specification
• continuity.py - Python CLI reference implementation
• examples/ - Usage guides and integration examples
• docs/ - Architecture and roadmap

Installation

# Clone repository
git clone https://github.com/[username]/continuity-protocol
cd continuity-protocol

# No dependencies required (uses Python stdlib only)
python continuity.py setup

# Initialize your first project
python continuity.py init "My Project" --goal "Your goal here"

Core Features

1. Session Tracking

python continuity.py session start
# ... work happens ...
python continuity.py session end --summary "What you accomplished"

2. Archaeological Reconstruction

python continuity.py archaeology --window 7

Shows chronological timeline of sessions, detects gaps >2 hours, presents evidence without interpretation.

3. Drift Detection

python continuity.py drift check

Checks alignment between stated goals and actual state. Alerts if:

• Goal undefined after 5+ sessions
• Gap >7 days since last session
• No snapshots after 10+ sessions

4. Specification Locks

python continuity.py lock core-workflow --version v1.0

Prevents proven patterns from drifting. Modification requires approval + version increment.

5. Clean Resumption

python continuity.py resume

Full context reconstruction: state, drift alerts, recent archaeology, locked specs.

Platform Adapters

The protocol can be implemented on existing platforms:

• Notion - Database + automation + slash commands
• GitHub - Issues + Actions + Project metadata
• Obsidian - Vault + Dataview + Templater
• Linear - Projects + Issues + API
• VS Code - Extension + workspace settings

See examples/integration_examples.md for implementation guides.

Design Principles

1. Artifact-first: Truth lives in stored artifacts, not memory
2. Human-governed: AI proposes, human decides
3. Evidence-based: Reconstruct from facts, not interpretation
4. Local-first: Works offline, zero dependencies, you own the data
5. Platform-agnostic: Protocol works anywhere with storage + compute

Use Cases

Who this serves:

• Solo founders juggling multiple product ideas
• Researchers with multiple concurrent projects
• Consultants context-switching between clients
• Creative technologists spawning experimental projects
• Anyone who sets projects down WIP and returns later

Works for:

• Portfolio pattern (multiple projects, frequent switching)
• Hyperfocus pattern (single project, sustained over months/years)

Documentation

• PROTOCOL.md - Full technical specification
• examples/basic_usage.md - Getting started guide
• examples/integration_examples.md - Platform adapters
• docs/ARCHITECTURE.md - How it works internally
• docs/ROADMAP.md - Development plan

Contributing

This is an open protocol. Contributions welcome:

• Protocol improvements (via issues/PRs)
• Platform adapters (link to your implementation)
• Bug reports and feature requests
• Documentation improvements

See ROADMAP.md for current priorities.

License

• Protocol specification: CC0 1.0 Universal (Public Domain)
• Reference implementation: MIT License

Use freely. Build implementations. Adapt to your needs.

Status

Current: v1.0 - Foundation complete

Next: v1.1 - Production polish (tests, multi-project, packaging)

See ROADMAP.md for full development plan.

---

The pattern is the value. Build on it.