360安全龙虾(Security Claw)
Agent security sandbox that isolates OpenClaw-style agents from the host system
What 360 Security Claw is
This address belongs to a 360 product page, and it delivered only an unrendered application shell when checked. The description here therefore rests on what can be read from the page's own metadata and bundled code rather than on marketing copy.
The keywords identify a product line called Nami Work, and the shipped code carries identifiers suggesting both a personal edition and an enterprise edition, so there is a real product behind the address even though the rendered page could not be read.
What it appears to do
The shell indicates a single-page web application for an AI work assistant, loading assets from a 360 content network and pulling shared libraries and behavioural analytics from the group's usual infrastructure. The implied loop is sign in, delegate work, review results.
Because the page rendered no body text, nothing more specific can be confirmed — no function list, no model list, no supported platforms and no pricing. The only reliable facts are the domain, the product naming found in the code, and the presence of separate personal and enterprise entry logic. Treat any capability claim about this product as unverified until you can load it yourself.
Who it is for
The metadata suggests individual knowledge workers and enterprise teams inside that vendor ecosystem looking for an AI assistant, with the two editions aimed at different sizes of buyer.
What to keep in mind
Do not base a decision on this listing. Pricing and features are unconfirmed, and the honest position is that the page could not be read rather than that it has nothing to offer.
Where an assistant connects to your documents and accounts, it runs with your authority, so the questions that matter are whether it processes locally or in the cloud, which directories and applications it may reach, and how long conversations are retained. None of those is answered on the page.
One detail worth knowing before typing anything confidential: the code shows analytics and monitoring libraries loading. That is ordinary for a web property, and it also confirms that usage signals leave the page, which is the kind of thing worth establishing before an assistant sees client material.
Confirm capabilities and terms on the rendered product, and decide what you are willing to let it reach rather than assuming the permissions are narrow. There is also a practical reason to revisit this entry later: a page returning only a shell may simply have been mid-deploy, and rendering behaviour varies with region, browser and time of day. If the product is relevant to you, ask the vendor for documentation directly, because a marketing page that fails to render is not the only source available.
Pros & cons
✓ What we like
- Product naming in the code points to a real, released product line
- Separate personal and enterprise entry paths are present
- Loads assets through the vendor's own content network
! What to watch out for
- The page rendered only an application shell, so nothing is confirmed
- No feature list, model list or pricing could be read
- Analytics and monitoring libraries load, and no retention terms are stated
FAQ
Are the features confirmed?
No. The page returned only a shell, so treat anything you read about the product as unverified until you load the rendered version.
Does it cost anything?
Pricing is not stated on the page. Confirm on the rendered product rather than assuming from this listing.
What should I check before using it?
Whether processing is local or cloud, which files and applications it can reach, and how long conversations are kept.
Last reviewed: 2026-09-19
More OpenClaw & desktop agents tools
View all →-
360龙虾卫士 Runtime guard that monitors AI agent file, process, network and screen activity Free tools Agent Workflow automation Openclaw skills Security -
ArkClaw Volcano Engine Doubao models Free tools Workflow automation API integration Multi model platform Openclaw skills -
AstronClaw iFLYTEK AI office assistant Open source tools Agent Workflow automation Developer tools Openclaw skills -
AutoClaw A general agent that researches, writes and executes tasks Free tools Agent Personal assistant Workflow automation Openclaw skills -
BoClaw Enterprise AI office platform Agent Workflow automation Developer tools Office productivity Openclaw skills -
CoPaw Portable agent runtime for on-prem or cloud deployment Free tools Agent Developer tools Model deployment Openclaw skills