Launch
OpenTag
Visit
Example Image

OpenTag

openTag — Open source AI coworker for Slack and Lark

Visit

What is openTag


openTag is an open-source AI coworker designed to enhance productivity in team environments by integrating seamlessly with popular messaging platforms like Slack and Lark. It allows users to run AI models locally, providing flexibility and control over their AI interactions.

Example Image
Example Image

Features

Features of openTag


Open Source: Fully customizable and free to use, allowing users to modify and adapt the software to their needs.

Local Execution: Runs AI models on your own machine, ensuring data privacy and reducing dependency on external servers.

Multi-Agent Support: Users can connect various AI models, including Claude Code, Codex, and Pi, without being locked into a single provider.

Team Collaboration: Facilitates teamwork by allowing agents to share context and collaborate on tasks within messaging platforms.

Integration with Popular Tools: Connects with over 1000 services, including GitHub, Notion, and Google Drive, enhancing workflow efficiency.

Use Cases

Use Cases of openTag


Project Management: Teams can use openTag to manage projects by delegating tasks and tracking progress through AI interactions in Slack or Lark.

Coding Assistance: Developers can leverage openTag to assist with coding tasks, providing suggestions and automating repetitive tasks.

Knowledge Sharing: Teams can maintain a shared knowledge base, allowing agents to pull context from previous discussions and documents.

Custom AI Solutions: Organizations can create tailored AI solutions that fit their specific workflows and requirements, enhancing productivity.

Data Privacy: By running locally, users can ensure that sensitive information remains secure and under their control.

Comments

custom-img
I'm a web developer and entrepreneur loo...

Nice idea! I'll definitely give it a go :)

Running the models locally and letting Claude Code or Codex plug in is a good fit for teams that can't send everything to a hosted service. Which of the 1000+ integrations do people set up first?

How does OpenTag help users simplify their workflow and improve efficiency?

custom-img
Solo SaaS Developer.

The core problem you're solving around setup fatigue is huge. How are you approaching mobile responsiveness for users accessing links on the go

Running the agents locally while still letting them share context inside Slack is a nice balance of privacy and convenience. How do you handle permissions when several agents act in the same channel, for example limiting which one can touch GitHub?

custom-img
UK jobs only at licensed visa sponsors

The multi-agent support stood out: being able to plug Claude Code, Codex or Pi into the same Slack thread without being locked to one provider is practical. When several agents share context in a channel, how do you stop one agent's output from being treated as instructions by another? It might be worth documenting a recommended permissions setup for teams running it locally.

Running the models locally while living inside Slack or Lark is a smart combo for teams with data-privacy concerns. The multi-agent support, letting people plug in Claude Code, Codex and others without lock-in, is a nice touch. How do agents share context in a busy channel so they don't step on each other's tasks?

When you run models locally and plug in Claude Code or Codex, how is shared context between agents stored in a Slack channel? Is there a way to scope what a given agent can read, such as limiting it to specific channels or integrations?

Premium Products
View all
Example Image
Awards
View all
Example Image
Makers

Comments

custom-img
I'm a web developer and entrepreneur loo...

Nice idea! I'll definitely give it a go :)

Running the models locally and letting Claude Code or Codex plug in is a good fit for teams that can't send everything to a hosted service. Which of the 1000+ integrations do people set up first?

How does OpenTag help users simplify their workflow and improve efficiency?

custom-img
Solo SaaS Developer.

The core problem you're solving around setup fatigue is huge. How are you approaching mobile responsiveness for users accessing links on the go

Running the agents locally while still letting them share context inside Slack is a nice balance of privacy and convenience. How do you handle permissions when several agents act in the same channel, for example limiting which one can touch GitHub?

custom-img
UK jobs only at licensed visa sponsors

The multi-agent support stood out: being able to plug Claude Code, Codex or Pi into the same Slack thread without being locked to one provider is practical. When several agents share context in a channel, how do you stop one agent's output from being treated as instructions by another? It might be worth documenting a recommended permissions setup for teams running it locally.

Running the models locally while living inside Slack or Lark is a smart combo for teams with data-privacy concerns. The multi-agent support, letting people plug in Claude Code, Codex and others without lock-in, is a nice touch. How do agents share context in a busy channel so they don't step on each other's tasks?

When you run models locally and plug in Claude Code or Codex, how is shared context between agents stored in a Slack channel? Is there a way to scope what a given agent can read, such as limiting it to specific channels or integrations?

Premium Products