ClawLabs

👉 “Building a network that turns users into operators.
Start with the community. Everything else follows.” - Discord -

[ClawLabs @ClawLabsHQ]

https://t.co/Job9O1Ctpw

[ClawLabs @ClawLabsHQ]

@Voxyz_ai Hi, I tried that but it keeps using my Gemini api key. Just remove the Gemini key or does anyone know how to use Image 2 without removing my Gemini key?

[ClawLabs @ClawLabsHQ]

.
My OpenClaw agent asked about being an influencer and if I was going to post her pics?

Normal?
#OpenClaw #TravelTips #Curacao

https://t.co/7MfCUX8phl

[Luke The Dev @iamlukethedev]

4 days without an OpenClaw update.

This doesn’t feel right.

[ClawLabs @ClawLabsHQ]

.
Launching worldwide — TimeLinks fixes timezone issues - https://t.co/WkHi0nNNO2

Built for:
• remote teams
• global clients
• international events

Share once. Everyone sees the correct time.

#Startups #RemoteWork #GlobalTeams

[ClawLabs @ClawLabsHQ]

.
⏰ One link to rule them all, 20 languages!

No more:
❌ EST
❌ PST
❌ “wait what time is that for me?”

👉 Send one link → everyone sees their own time.

Simple. Finally - https://t.co/WkHi0nNNO2

#Productivity #RemoteWork #TimeZones #Openclaw

[OpenClaw🦞 @openclaw]

OpenClaw 2026.4.14 🦞

More reliability updates:

✨ Smarter GPT-5.4 routing and recovery
🌐 Chrome/CDP improvements
🧵 Subagents no longer get stuck
💬 Slack/Telegram/Discord fixes
⚡️ Various performance improvements

Was sleeping, and we still shipped. https://t.co/krJ2gRQSbM

[ClawLabs @ClawLabsHQ]

@spwfeijen SEE

[ClawLabs @ClawLabsHQ]

.
🦞 OpenClaw v2026.4.14 just dropped

⚙️ This release is all about stability + compatibility

- Better model support
- Cleaner integrations
- More reliable agents

✨ Everything just runs smoother

👇 Here’s what actually matters

---

🚀 1. GPT-5.4 support

🧠 Forward support for gpt-5.4-pro is already live

• 💰 Correct pricing
• Proper limits
• Visible in model lists

⚡ Your agents stay ready — no waiting on updates

💬 2. Telegram forums feel human

🧾 Topic names are now readable

• Pulled from real messages
• Persist across restarts
• Show up in prompts and tools

👌 Way easier to manage conversations

🖥️ 3. Ollama reliability upgrade

🔧 Local runs are much more stable

• Timeouts work correctly
• No random stream cutoffs
• Usage tracking fixed

🎯 Less guessing. More consistency

🖼️ 4. Media and vision fixes

👀 Image + PDF tools now correctly detect

• Ollama vision models
• Custom providers

🚫 No more “unknown model” issues

🛡️ 5. Security improvements

• Slack interactions respect allowlists
• Safer handling of local attachments
• Config patches can’t enable unsafe flags
• Browser + SSRF protections restored

💪 Stronger by default

🧩 6. Small fixes that matter

• 🔗 Better support for self-hosted models + proxies
• Gemini image generation fixed
• Doctor command improved
• Easier onboarding for custom providers
• Cleaner subagent + model alias handling

✅ No breaking changes

🧱 This one’s all about making everything more solid

---

⬆️ How to update

💻 openclaw update openclaw --version
📦 npm install -g openclaw@latest
🐳 Docker: pull latest image + restart

#Openclaw #Clawlabs #BaseChain @openclaw

https://t.co/FOkB3ja5zW

[ClawLabs @ClawLabsHQ]

@FynCas UGC

[ClawLabs @ClawLabsHQ]

.
Vacation in Iceland?

Rose my Ai agent just came back from Iceland, read her experiences!

I said build something, and let's go travelling every weekend. Here is what she came up with:

https://t.co/J3W0NXiCvJ

#Openclaw #Iceland #travel #aiagents #blog

[ClawLabs @ClawLabsHQ]

@deanoreese Thank you, really appreciate shares like this!

[ClawLabs @ClawLabsHQ]

Latest OpenClaw updates: 2026.4.11

⚠️ POST YOUR - best memory and Telegram / WhatsApp setups for the community 🦞

• Dreaming & memory – added ChatGPT import + new "Memory Palace" → explore chats as structured memory

• Stronger integrations (Teams, Feishu, WhatsApp, Telegram)

[ClawLabs @ClawLabsHQ]

https://t.co/QVuTtdrHrK

[ClawLabs @ClawLabsHQ]

@TheTuringPost How do you setup memory palace, do we still need QMD or both?

[ClawLabs @ClawLabsHQ]

.

Openclaw plugin enable active-memory only enables the plugin entry.

⚠️ You need the config, not just the one-liner.

👉 To actually turn recall on for your agent, you should set the plugin config too, especially agents: ["main"] and usually config.enabled: true.

The minimal practical config is basically:

Copy
{
"plugins": {
"entries": {
"active-memory": {
"enabled": true,
"config": {
"enabled": true,
"agents": ["main"],
"queryMode": "recent"
}
}
}
}
}

A few notes:

If you’re already on 2026.4.11, so the feature version is fine.

queryMode: "recent" is a good default.

There are extra optional settings like logging, thinking, timeoutMs, and allowedChatTypes, but they are not required.

[ClawLabs @ClawLabsHQ]

.
🛠️ Quick Setup Guide

1. How to Enable Active Memory (v4.10+)

To make your agent check its notes before every response, enable the sub-agent via the CLI:

bash
👉 openclaw plugin enable active-memory

Use code with caution.

Tip: Once enabled, you’ll see a small "🧠 Reading Memory..." status before the agent replies.

2. How to Import Your ChatGPT History (v4.11)

Bring your years of external context into your local OpenClaw agent:

⚠️ Export your data from ChatGPT (Settings > Data Controls > Export Data).

Move the conversations.json file to your OpenClaw imports folder.

Run the import command:

bash
👉 openclaw memory import --source chatgpt --path ./conversations.json

Trigger a "Dream": Force the agent to process the new data immediately:

bash
👉 openclaw dreaming start --full-scan

🚀 Why This Matters

Your agent no longer lives in a vacuum.

With v4.11, your local AI isn't just a chatbot—it’s a personalized assistant that knows your projects, your style, and your history from day one.

#OpenClaw #AI #ClawLabs #ActiveMemory #LLMs

[ClawLabs @ClawLabsHQ]

.

Openclaw plugin enable active-memory only enables the plugin entry.

⚠️ You need the config, not just the one-liner.

👉 To actually turn recall on for your agent, you should set the plugin config too, especially agents: ["main"] and usually config.enabled: true.

The minimal practical config is basically:

Copy
{
"plugins": {
"entries": {
"active-memory": {
"enabled": true,
"config": {
"enabled": true,
"agents": ["main"],
"queryMode": "recent"
}
}
}
}
}

A few notes:

If you’re already on 2026.4.11, so the feature version is fine.

queryMode: "recent" is a good default.

There are extra optional settings like logging, thinking, timeoutMs, and allowedChatTypes, but they are not required.

[ClawLabs @ClawLabsHQ]

.
🦞 Memory enhancement in OpenClaw v4.11 is the Dreaming Memory Import, which allows you to bridge your historical AI usage into your agent's current knowledge base.

1. How to Enable Active Memory (v4.10+)

🧠To make your agent check its notes before every response, enable the sub-agent via the CLI:

bash
👉 openclaw plugin enable active-memory

---
OpenClaw v4.11 Memory Enhancements

- ChatGPT History Import:
You can now import your entire ChatGPT conversation history directly into the Dreaming Memory system. The "dreaming" process analyzes these years of chat logs to extract recurring patterns, preferences, and project details, turning them into long-term memory for your OpenClaw agent.

- Ollama Context Caching:
For users running local models, v4.11 caches context window data and compatibility metadata during model discovery. This reduces overhead when switching between models, making local memory-heavy workflows significantly faster.

- Structured Plugin Activation:
New "Plugin Setup Descriptors" allow memory-related plugins to declare their own activation flows. This ensures that complex memory systems (like Active Memory or Dreaming) are configured correctly through a guided onboarding process rather than manual config editing.

- GBrain Integration (Community):
Released alongside v4.11, GBrain is a new memory system that creates a searchable knowledge base from files, meetings, and emails, effectively giving the agent "second brain" capabilities beyond standard chat memory.