Documentation

Documentation

Welcome to the Conduct documentation. Learn how to manage AI agents at 100x speed with specification-driven development.

Quick Navigation

Getting Started
Install and initialize Conduct in minutes
 CLI Reference
Complete command reference
 AI Agents
Integration with AI coding tools
 Architecture
How Conduct works under the hood

What is Conduct?

Conduct v0.1 is an AI agent orchestration platform with persistent memory that works seamlessly with 7+ AI coding assistants. It provides a structured spec-run-check workflow with a libSQL memory database that remembers all your work:

  • Cursor - Custom command integration (9 templates)
  • Claude - Automatic context loading via CLAUDE.md
  • Warp - Agent Mode command files
  • Gemini - Instructions via GEMINI.md
  • Factory - Workflow integration
  • OpenCode - Command system integration
  • GitHub Copilot - Prompt files for context

Key Features

Persistent Memory Database

All work stored in libSQL (SQLite-compatible) with local or remote modes:

conduct init
# Choose: Local (embedded) or Remote (Turso)
# Memory persists: specs, runs, checks, features, issue connections

What’s Remembered:

  • ✅ Features discovered in your codebase
  • ✅ Specifications with versions and status
  • ✅ Execution runs with feature links
  • ✅ Verification checks and results
  • ✅ Issue tracker connections (GitHub, Linear)

Spec → Run → Check Workflow

Three-phase workflow with memory integration:

1. Specification - What to build

# AI agent uses conduct-spec template
→ Creates: _conduct/specs/1.v0.spec.md
→ Registers in memory database

2. Run - How it was built

# AI agent uses conduct-run template
→ Creates: _conduct/runs/1.v0.run.md
→ Links to discovered features
→ Logs execution to memory

3. Check - Verification

# AI agent uses conduct-check template
→ Creates: _conduct/checks/1.v0.check.md
→ Records pass/fail results

11 CLI Commands

Complete memory management:

conduct init        # Initialize project
conduct discover    # Auto-discover features
conduct list        # Query memory database
conduct reconcile   # Sync memory with codebase
conduct health      # System health check
conduct config      # Manage configuration & trackers
conduct upgrade     # Update templates & instructions
# + save, sync, relevancy, archive

Directory Structure 

_conduct/
├── specs/              # Specifications
├── runs/               # Execution logs  
├── checks/             # Verification reports
├── designs/            # UI designs (optional)
├── operations/         # SQL files for memory
├── templates/          # 9 agent command templates
├── memory.db           # libSQL database
└── .meta/              # AI-generated docs

conduct.json            # Configuration with JSON schema
conduct.track.json      # Fast local tracker
AGENTS.md               # Symlink to _conduct/AGENTS.md
CLAUDE.md               # Claude instructions
WARP.md                 # Warp instructions
GEMINI.md               # Gemini instructions

Issue Tracker Integration

Configure and sync with external trackers:

conduct config tracker add
# Select: GitHub or Linear
# Auto-link specs to issues
# AI agents reference issue context

Supported:

  • GitHub Issues
  • Linear
  • Jira (planned)
  • GitLab (planned)
  • Azure DevOps (planned)

JSON Schema Support

IDE autocomplete and validation:

  • conduct.json: https://docs.conduct.run/schema/conduct.json
  • conduct.track.json: https://docs.conduct.run/schema/track.json

Add $schema property for instant IntelliSense!

Next Steps

New to Conduct? Start here:

  1. Install and initialize Conduct in your project
  2. Learn about AI agent integration
  3. Explore the CLI commands
  4. Understand the architecture
  5. Read the comparison with other tools
Getting Started