Agent access through MCP
Connect an existing coding or chat agent to Kody over MCP, so the agent can discover and invoke the capabilities available to the signed-in account.
Kody
Kody is a hosted execution and automation layer that connects AI coding and chat agents to saved capabilities, credentials, integrations, and reusable packages. It helps developers turn one-off agent tasks into scheduled jobs, webhooks, workflows, and hosted tools.
A persistent execution layer for AI agents
Kody is a hosted factory for agent-driven capabilities. It connects to the AI agents a user already works with over MCP, giving those agents two core tools: search for available capabilities and execute TypeScript code on Kody's servers.
A task can start as a one-off conversation and become durable software. Users can save behavior as versioned packages, attach state and credentials, and trigger it through schedules, webhooks, authenticated HTTP requests, workflows, or hosted apps. Kody is designed to keep the agent responsible for reasoning while providing a persistent execution layer that can be reused across agents and kept running when a laptop is closed.
Core capabilities
Capability discovery
Use search to find saved packages, connected services, official guides, relevant memories, and available capabilities before building a workflow.
Server-side execution
Run temporary TypeScript modules on Kody's servers that compose discovered capabilities, call connected services, and import exports from owned packages.
Reusable packages
Turn useful behavior into named, versioned packages with repository-backed source and isolated durable packageStorage() for runtime state such as cursors, preferences, and checkpoints.
Jobs, workflows, and hosted surfaces
Run package code through recurring schedules, deferred workflows, authenticated HTTP calls, webhooks, or package-hosted browser and HTTP surfaces.
Credentials, integrations, and memory
Store secrets, OAuth-backed integrations, remote MCP connections, and account memories separately so credentials stay server-side and useful context can persist between agents and conversations.
Practical uses
Track repository activity
Ask an agent to inspect GitHub activity using a saved credential, filter meaningful events such as releases or new repositories, and save the result as an export for faster reuse later.
Automate engineering checks
Create a scheduled package that checks CI results overnight and routes likely flaky failures to an agent for investigation, without requiring the developer's laptop to stay open.
Respond to service events
Receive an event from a service such as Sentry through a webhook, fingerprint repeated issues, and trigger an agent-assisted triage workflow.
Handle recurring business tasks
Build an email- or purchase-triggered workflow that uses saved context to route messages or draft a contextual response through connected services.
Create personal operational tools
Expose a small package-hosted application or authenticated HTTP endpoint for tasks such as querying account data or performing an approved operation through a connected integration.
Pros and Cons
Pros
- Works with multiple coding and chat agents through MCP, reducing dependence on a single agent host.
- Lets users move from temporary tasks to versioned, reusable packages with durable runtime state.
- Keeps saved secrets and OAuth tokens server-side instead of returning credential values to the assistant.
- Supports scheduled jobs, webhooks, workflows, authenticated HTTP calls, and hosted package surfaces.
- Open-source code can be read, forked, and self-hosted according to the site.
Cons
- The hosted environment cannot directly reach local disks, localhost, local-only processes, or devices on a private home network; those resources need a protected public HTTPS route and remote MCP connection.
- Work involving large npm graphs, native code, or multi-minute CPU may not fit a Worker isolate and requires an external process operated by the owner.
- Team and enterprise offerings are described as an evolving contact-sales option rather than a fully detailed public plan.
FAQ
How does Kody work with an AI agent?
Kody connects to an agent over MCP. The agent can use search to discover capabilities, integrations, packages, guides, and memories, then use execute to run a temporary TypeScript module on Kody's servers.
Can a one-time agent task become reusable software?
You can use a one-off execute run for exploration, or turn useful behavior into a named, versioned package stored in a repository. Packages can expose exports, apps, webhooks, and scheduled jobs.
How are credentials handled?
Kody stores credentials privately and lets runtime code refer to them through placeholders or approved package mounts. The saved secret value is not returned to the assistant, and OAuth integration tokens remain server-side.
Can Kody access local files or services?
Kody's hosted environment cannot directly access a Mac's disk, local-only processes, localhost, or devices limited to a home network. A local capability must be exposed through a protected public HTTPS route and connected as a remote MCP server.
What plans does Kody offer?
The Free plan is $0 and includes the whole factory with limits including 5 jobs and a fastest job interval of 15 minutes. Standard and Pro are paid monthly plans, with annual billing rates shown on the pricing page; teams and enterprise users are invited to contact Kody.
Quick Facts
- Category
- Developer tool / AI agent infrastructure
- Connection model
- MCP connection to the user's AI agent
- Execution
- Temporary TypeScript modules run on Kody's servers
- Persistence
- Versioned packages, package storage, memories, jobs, and workflows
- Integrations
- Connected services, OAuth integrations, remote MCP servers, webhooks, and package apps
- Plans
- Free, Standard, Pro, plus contact-sales Teams / Enterprise
Alternative a Kody
CreateOS Sandbox
CreateOS Sandbox is an isolated compute environment for running code and agent workloads inside Firecracker micro-VMs. It is designed for workflows that need machine-level isolation, private networking between sandboxes, and programmatic control through SDK, CLI, or MCP.
AakarDev AI
AakarDev AI helps teams manage AI provider access, project-level setups, logs, and analytics from one dashboard. It supports BYOK workflows and lists providers including OpenAI, Google Gemini, Anthropic, Groq, Mistral AI, and Perplexity AI.
Trigger.dev chat agent
Trigger.dev chat agent is a durable AI chat backend for developers building stateful conversations that can survive refreshes, crashes, and long-running turns. It connects with the AI SDK `useChat` flow and runs on managed infrastructure with no timeout on a turn.
Arduino VENTUNO Q
Arduino VENTUNO Q is an edge AI computer for AI and robotics applications. It combines AI inference and deterministic control on a single board and is designed to work with Arduino App Lab.
ByteAsk
ByteAsk is a terminal-first AI coding agent for C and C++ that edits repositories and verifies changes with the real compiler, debugger, sanitizers, and tests before showing a diff. It offers a free tier plus paid plans, with editor connectors and zero-retention handling described in the source.
Codex Plugins
Codex Plugins bundle reusable skills, app integrations, and MCP servers into workflows you can install in the Codex app or use from Codex CLI. They help extend Codex with connected-service tasks, reusable instructions, and shared team workflows.