Find apps and tools you can use right away
Search, inspect, and run before building from scratch.
One CLI lets agents use the apps you already use
Let Codex, OpenClaw, Claude Code, and other agents use your apps, APIs, and internal tools. Then use OOMOL Studio to compose and extend your own tools.
Start with the apps you already use, then bring in your own tools
SaaS apps, internal services, and custom APIs can all follow the same path into agents.
GitHub
Slack
Notion
Gmail
Linear
Your API
Bring business endpoints and internal services onto the same path.
Your logic
Compose and extend your own implementation when ready-made tools stop short.
What agents need to run for real
Get one tool working first, then extend it into your own toolchain.
Connect accounts and API keys only when a tool really needs access
Keep OAuth, API keys, and access relationships on the same path without putting trust cost first.
When ready-made tools stop short, bring in your APIs and business logic
Add internal services, custom endpoints, and composed logic onto the same path.
Carry the same implementation into different surfaces
Let CLI, automations, and other agent entry points share one implementation without repackaging it.
Once one tool is connected, agents can do things like this
Start with outcomes, then explain how it connects.
GitHubSlack
Send a PR summary to Slack
Turn code changes and team communication into one action.
You can say
Summarize this PR and send it to Slack.
The agent will do
Read the diff, write the summary, post it to the target channel.
NotionLinear
Turn a Notion spec into Linear
Keep docs, breakdown, and tracking in one chain.
You can say
Read this Notion page, break it into tasks, and sync them to Linear.
The agent will do
Read the doc, split the tasks, write them back to Linear.
GmailYour API
Read a Gmail attachment, then call your API
Put ready-made input and custom processing on one path.
You can say
Read the Gmail attachment, call our PDF API, and return the result.
The agent will do
Take the email attachment, then keep going through your API.
Start in oo-cli. Build your own only when ready-made tools stop short
Get one tool working first. When ready-made tools are not enough, build your own in Studio, then hand it to Cloud when it needs to keep running and be delivered.
Install oo-cli and let agents call apps and tools
In Codex, OpenClaw, Claude Code, and terminal workflows, start by searching, inspecting, and running published tools to get the usage path working.
Agent demo video
Show installing, searching, inspecting, and running a tool in Codex.
When ready-made tools are not enough, build your own in OOMOL Studio
Tell the agent what you want, then keep editing code, dependencies, parameters, and compositions yourself. Studio does not replace engineering workflow; it helps you create new tools faster and validate them locally.
Studio demo video
Show the path from prompting to generating a tool and validating it locally.
When it needs to keep running and be delivered, Cloud takes over
After local validation is done, Cloud handles runtime, configuration, secrets, and delivery relationships. You do not need to rebuild another delivery layer around the same implementation.
Handle runtime and delivery in one path
Keep the same implementation as you deliver the tool to yourself, your team, or customers.
Keep configuration and usage relationships in one backend
Manage secrets, access, releases, runtime settings, and usage data in one place.
Cloud console preview
Bring runtime settings, secrets, and delivery relationships into one backend.
Start with ready-made tools, build your own only when needed
Use oo-cli to get ready-made tools working first. When you need custom ones, move into Studio and Cloud to compose, extend, and deliver your own tools.