Identity
Proof-backed identity
Register the agent and the approving account as trust receipts any connected tool can check.
Your Agents Have the Keys
Trusted identity, durable context, and authority that can be revoked. Checked before they act.
What agents get
Six capabilities, backed by receipts
Identity, permissions, memory, jobs, and portable checks. Soft atmospheric proof surfaces, not another chat wrapper.
Permissions
Scoped authority
Record allowed actions, spending caps, expiration, and revocation. Set them once, then reuse.
Before action
Limits checked first
Connected runtimes read the current receipt and stop unsupported actions before they run.
Memory
Receipt-backed context
Durable work becomes reusable evidence, so agents keep continuity without starting from zero.
Jobs
Scheduled with evidence
Run recurring work with a clear trail: who authorized it, what ran, and what was recorded.
Everywhere
One setup, many tools
The same identity and permissions follow across Cursor, Claude Code, Codex, and your stack.
Before the action
Watch the runtime stop the action
Connected runtimes check who the agent is, what it may do and spend, then stop anything outside those limits before it runs.
~/neus
Ready$npx @neus/sdk setup
Cursor connected
Claude Code connected
Signed in | profile ready
$neus check
Identity: current
Permissions: scoped
Trust ready before action
Framework-neutral
Keep your stack. Add authority.
Frameworks run the tools. NEUS checks identity, context, and authority before they run. Agent permissions · Receipt format.
NEUS | reusable receipt
Portable
One proof. Every surface that accepts it.
- First useVerify once
- Next appReuse
- Agent actionReuse
- Network effectCompounds
Profile owners control visibility. Connected systems check status.
Agent & assistant FAQ
Identity, permissions, and receipt checks before sensitive actions.
Ready to add control?
One setup. Same identity, context, and limits in Cursor, Claude Code, Codex, and production.