Introduce typed tool plugins and CLI build tools
TL;DR
Added `defineToolPlugin` and new CLI commands (`build`, `validate`, `init`) to streamline the creation of typed tool plugins with generated manifest metadata and context factories.
What changed
OpenClaw added the defineToolPlugin helper along with three new CLI commands: init, validate, and build. The changes generate manifest metadata and typed context factories automatically during plugin creation.
The update targets developers who extend the agent with custom capabilities. It replaces manual manifest writing and loose typing with a structured workflow that produces consistent plugin output.
Why it matters
Typed plugins reduce runtime errors when the agent loads community or personal skills from ClawHub. This matters for Vibe Builders who maintain their own automation surface instead of relying on closed cloud agents.
The bet is that more reliable plugin development will increase the number of safe, shareable skills without forcing every user to audit code. It directly competes with less structured agent runtimes that leave extension work to ad-hoc scripts.
How to use it
Run npx openclaw init inside a new directory to scaffold a typed plugin. Edit the generated TypeScript file that defineToolPlugin exports, then execute openclaw validate followed by openclaw build.
The resulting manifest and context factory files are ready for local testing or submission to ClawHub. No additional paid plan is required; the commands are part of the free MIT-licensed CLI.
Watch for
Confirm the bet if published plugins show fewer type-related failures in agent logs over the next month. The approach breaks if context factories cannot express complex state that some browser or shell tasks require. Expect a follow-up release that adds hot reload for faster iteration during development.
Who this matters for
- Vibe Builders: Use the new CLI to scaffold and build custom agent skills without writing manual manifest files.
- Developers: Implement defineToolPlugin to automate manifest generation and ensure type safety for agent extensions.
Harsh’s take
OpenClaw is moving toward a more professionalized developer experience by enforcing structure on what used to be ad hoc scripts. The introduction of a dedicated CLI for building and validating plugins suggests they are prioritizing ecosystem stability over raw speed. By automating manifest metadata and context factories, they reduce the friction of contributing to ClawHub while preventing the runtime crashes that plague less structured agent frameworks.
This update signals that the agent space is maturing beyond simple wrappers. Forcing developers into a typed workflow is a smart move to ensure that community-contributed skills actually work when deployed. If you are building custom automation surfaces, adopting these typed plugins is the only way to maintain a reliable system as your agent's capabilities scale.
by Harsh Desai
About OpenClaw
View the full OpenClaw page →All OpenClaw updatesGo deeper
More AI news
- Daily RoundupGLM 5.3 and Vercel agent tools land, Gemini SAT practice, plus fresh models on Hugging Face
Google and Vercel released new agent and model features while Hugging Face and Replicate added trending models; OpenAI tightened safety after agent issues.
- Daily RoundupVercel-Origin beta and GPT-5.6 Sol discount, plus agent tools for builders
Vercel rolled out Origin repository deployments and a month-long GPT-5.6 Sol discount on AI Gateway while Product Hunt featured new agent layers and Hugging Face highlighted an uncensored Qwen model, with industry moves around data use and automation exits.
- Weekly DigestHermes Agent v0.20 voice and A2A protocol, OpenClaw mobile updates, and multi-agent tools to test now
Hermes Agent rolled out v0.20 releases with voice streaming, A2A interoperability, desktop artifacts, and self-recovery tools while OpenClaw added Android switching, iOS caching, and plugin provenance checks across the week.