Conversation
Code Review SummaryStatus: 2 Issues Found | Recommendation: Address before merge Overview
Fix these issues in Kilo Cloud Issue Details (click to expand)CRITICAL
WARNING
SUGGESTION
Other Observations (not in diff)Issues found in unchanged code that cannot receive inline comments:
Files Reviewed (12 files)
|
Co-authored-by: kiloconnect[bot] <240665456+kiloconnect[bot]@users.noreply.github.com>
|
The latest updates on your projects. Learn more about Vercel for GitHub.
|
pandemicsyn
left a comment
There was a problem hiding this comment.
lgtm, mostly small nit picky stuff you can ignore. Only thing we really need to update is the instance info.
| 6. Click **Create & Provision** | ||
|
|
||
| Your instance will be provisioned and ready within seconds. | ||
| Your instance will be provisioned in seconds. Each instance runs on a dedicated machine with 2 shared vCPUs, 4 GB RAM, and a 20 GB persistent SSD. Once created in a region, your instance always runs there. |
There was a problem hiding this comment.
| Your instance will be provisioned in seconds. Each instance runs on a dedicated machine with 2 shared vCPUs, 4 GB RAM, and a 20 GB persistent SSD. Once created in a region, your instance always runs there. | |
| Your instance will be provisioned in seconds. Each instance runs on a dedicated machine with 2 shared vCPUs, 3 GB RAM, and a 10 GB persistent SSD. Once created in a region, your instance always runs there. |
| - **Restart** — Stop and start the instance | ||
| - **Start Machine** — Boot a stopped instance (up to 60 seconds) | ||
| - **Restart OpenClaw** — Quick restart of just the OpenClaw process; the machine stays up | ||
| - **Redeploy** — Update and restart the whole machine; all your files and data are preserved |
There was a problem hiding this comment.
| - **Redeploy** — Update and restart the whole machine; all your files and data are preserved | |
| - **Redeploy** — This will stop the machine, apply any pending image or config updates, and restart it. The machine will be briefly offline. |
| - **Redeploy suggested** — Redeploy is optional and only needed if you want to use the new behavior | ||
|
|
||
| For example, if you manually configured a channel such as Telegram, and would prefer to have KiloClaw manage the channel for you, you would need to redeploy. | ||
| When someone messages your bot for the first time, or a new device connects to the Control UI, you'll see a pairing request on the dashboard that you need to approve. See [Pairing Requests](/docs/automate/kiloclaw/chat-platforms#pairing-requests) for details. |
There was a problem hiding this comment.
| When someone messages your bot for the first time, or a new device connects to the Control UI, you'll see a pairing request on the dashboard that you need to approve. See [Pairing Requests](/docs/automate/kiloclaw/chat-platforms#pairing-requests) for details. | |
| When you initialize a new channel for the first time, or a new device connects to the Control UI, you'll see a pairing request on the dashboard that you need to approve. See [Pairing Requests](/docs/automate/kiloclaw/chat-platforms#pairing-requests) for details. |
Not sure that its always message driven so hedging bets a bit.
|
|
||
| ### When should I use Restart OpenClaw vs Redeploy? | ||
|
|
||
| - **Restart OpenClaw** — Restarts just the OpenClaw process. The machine stays up. Use this for quick recovery from a process-level issue. |
There was a problem hiding this comment.
| - **Restart OpenClaw** — Restarts just the OpenClaw process. The machine stays up. Use this for quick recovery from a process-level issue. | |
| - **Restart OpenClaw** — Restarts just the OpenClaw process. The machine stays up. Use this for quick recovery from a process-level issue or when you want to apply openclaw config changes. |
| ### I changed the model but the agent is still using the old one | ||
|
|
||
| After selecting a new model, click **Save & Provision** to apply it. This refreshes the API key and saves the new model. You may also need to **Restart OpenClaw** for the change to take full effect. |
There was a problem hiding this comment.
Don't have to change this now. But was also chatting with John today - when we native gateway we should probably just default to auto or something and push users towards the OpenClaw Control UI to change model defaults and the like. Because chances are thats where most folks will go to manage model+agent pairings and the like anyway.
| ### Safe Bins | ||
|
|
||
| A small set of stdin-only binaries (`jq`, `cut`, `uniq`, `head`, `tail`, `tr`, `wc`) are allowed in `allowlist` mode without explicit entries. These enforce strict stdin-only behavior — positional file arguments, path-like tokens, shell expansion, and redirections are all rejected. |
There was a problem hiding this comment.
| ### Safe Bins | |
| A small set of stdin-only binaries (`jq`, `cut`, `uniq`, `head`, `tail`, `tr`, `wc`) are allowed in `allowlist` mode without explicit entries. These enforce strict stdin-only behavior — positional file arguments, path-like tokens, shell expansion, and redirections are all rejected. |
I actually still get prompted for these the first time, so might be worth just ommiting this for now until we can poke at it some more.
Context
this provides a much more comprehensive overview of Kilo Claw and the current state of how it works so that users can understand more of the technical details about the instances that are running on and how that interacts with OpenClaw.
That is accomplished through starting to break the Kilo Claw documentation into a number of pages: