KendaliAI (Go Edition)

KendaliAI is a self-hosted, autonomous AI coding agent and orchestration gateway rebuilt natively in Go. Influenced heavily by actor-inspired systems and microkernel architectures, KendaliAI uses dynamic tool invocation, recursive LLM cognition loops, resource locking, and process supervision to execute local filesystem actions.

---

1. Quickstart

Prerequisites

• Go 1.20+
• SQLite3
• CGO (required by go-sqlite3)

1. Initialize Configuration

Generate a commented default config.yaml configuration file:

go run ./cmd/kendaliai config init

This creates ./config.yaml at the root directory:

version: 1

database:
  path: ./build/kendaliai.db

defaultProvider: deepseek

chatProviders:
  - name: deepseek
    type: deepseek
    apiKey: ${DEEPSEEK_API_KEY}
    model: deepseek-chat
  - name: openai
    type: openai
    apiKey: ${OPENAI_API_KEY}
    model: gpt-4o

embedding:
  apiKey: ${OPENAI_API_KEY}
  endpoint: https://api.openai.com/v1
  model: text-embedding-3-small

channels:
  - id: telegram-main
    channelName: telegram
    channelType: telegram
    token: ${TELEGRAM_TOKEN}

2. Validate Setup

Verify syntax and environment diagnostics:

# Validate configuration parameters
go run ./cmd/kendaliai config validate

# Verify local dependencies (Go, Git, SQLite, Permissions)
go run ./cmd/kendaliai doctor

---

2. CLI Command Index

Operate the microkernel daemon using the unified command suite:

Command	Description
kendaliai start	Start gateway process in foreground.
kendaliai start -d	Start gateway background daemon.
kendaliai stop	Stop background daemon.
kendaliai restart	Restart daemon.
kendaliai status	Show detailed uptime, CPU, memory, active agents, and cost metrics.
kendaliai logs	Stream system logs (--follow, --agent, --session, --level, --json).
kendaliai doctor	Diagnose platform dependencies and configurations.
kendaliai dashboard	Launch local TUI dashboard.
kendaliai config show	Dump active configuration.

---

3. Platform Architecture

                                 User Channel / API Gateway
                                             │
                                             ▼
                                  Conversation Engine
                                             │
                                             ▼
                                     Goal Tree Engine
                                             │
                                             ▼
                                     Planner / Reasoner
                                             │
                                             ▼
                                      Execution Graph
                                             │
                                             ▼
                                     Execution Scheduler
                                             │
                                             ▼
                                      Supervisor Tree
                                             │
                                             ▼
                                      Agent Processes
                                             │
                                             ▼
                                    Capability Runtime
                                             │
                                             ▼
                                      Executor Registry
                                             │
                                             ▼
                                     Runtime Environment
                                             │
                                             ▼
                                    Target Workspaces

   ────────────────────────────────────────────────────────────────────────────
    Event Store • Projection Engine • Telemetry & Tracing • Memory Broker Bus

---

4. Custom Skills

KendaliAI supports custom skills located in ~/.gemini/config/skills/ (global) or .agents/skills/ (workspace): Add specialized instruction guidelines by creating a Markdown file with YAML frontmatter. These are registered as tools loaded on demand:

Example: .agents/skills/frontend-design/SKILL.md

---
name: frontend-design
description: Create distinctive, production-grade frontend interfaces.
---
## Principles
- Use Outfit/Roboto fonts.
- Avoid generic colors.
- Use smooth gradients.

---

5. Verification & Tests

To run the full suite of MAK OS primitives integration tests:

go run ./tests/test_mak_run.go