← Back to Playbooks
Security

Harness: Permission Outside the Agent

The agent proposes; a host authority grants connectors, egress, and spend

Operator recipe from Meta’s Muse Secure VM pattern: keep permission outside the model. Sentinel (or an equivalent host authority) is the sole grantor of connector actions and network egress. The agent proposes; it does not approve. Credentials live in a vault; the runtime sees only surrogates until the host inserts the real token at an authorized boundary. Human approvals are scoped capabilities (one use, session, task, window, or standing grant) — not a chat “yes.” Treat behavioral evals and egress allowlists as acceptance tests before widening autonomy. Purchase path: single-use scoped cards (Muse partners with Stripe Link at launch; Stripe’s agent wallet issues a one-time card or shared payment token after a spend request). Primaries: Meta AI Muse security blog; Meta “Introducing Muse” (about.fb.com / Meta Newsroom); Stripe “Giving agents the ability to pay.” Related catalog tools: Stripe Link, METR Inspect Hawk, OpenAI Agents SDK (in-app harness, not the managed Agents API).

0
Registrations
$1,000
Prize Pool
Feb 23, 2026
Starts
0%
Complete

Core Workflows (6)

Each workflow represents a critical business function. Click any workflow to see detailed automation architecture.

01

Host authority is the sole grantor

Led by: Host Authority (Sentinel)

Run a Sentinel-class host-side authority separate from the agent cell. It alone grants connector methods and all network egress. Muse: the agent proposes; only Sentinel can grant. Replicate that split even if you are not on Muse — the OpenAI Agents SDK is an in-app harness you host; it is not OpenAI’s managed Agents API.

Sub-Agents
Proposing Agent
Skills Required
Muse Secure VMOpenAI Agents SDK
Human TouchpointConfirm the model cannot grant its own connectors or egress
02

Agent proposes; host evaluates

Led by: Proposing Agent

Connector calls submit connector, method, action class, scope, and user-task context to the host. The host evaluates policy and allows, denies, or asks the human. Egress is inspected at L4 and L7 (hostname, resolved IP, port, protocol, method, path, decoded request), including SSRF protections so a public name cannot resolve to private infrastructure.

Sub-Agents
Host Authority (Sentinel)
Skills Required
Connector policyEgress proxy
Human TouchpointPolicy for allow / deny / ask is set by the operator, not the model
03

Vault plus surrogate tokens

Led by: Host Authority (Sentinel)

Real secrets stay in a vault/auth daemon. Runtime code sees surrogate tokens. After the host authorizes a concrete request, it replaces surrogates with the real credential at the network boundary. Workers get explicit credential allowlists (a calendar worker cannot fetch an email token by rewriting a parameter). Prompt-injection cannot reveal a secret the agent never saw.

Sub-Agents
Credential Vault
Skills Required
authd / vaultSurrogate tokens
Human TouchpointNever put live PANs, OAuth refresh tokens, or API keys in the agent context
04

Human approvals as scoped capabilities

Led by: Approval Operator

When the host asks the user, the prompt is a client dialog describing the exact action — not a free-form chat reply the model can reinterpret. Grants can be one-use, session, task, time-window, or standing, and later requests must stay inside that scope. Tainted egress (process read untrusted/user data) drops auto-allow and falls back to approval.

Sub-Agents
Host Authority (Sentinel)
Skills Required
Scoped grantsTainted-egress policy
Human TouchpointApprove the scoped capability in the host UI, not inside the agent transcript
05

Evals and egress allowlists as acceptance tests

Led by: Eval Operator

Before widening auto-allow, run behavioral evals (prompt-injection / tool-abuse suites) and keep an egress allowlist. Muse uses model-level evals plus agentic red-teaming; METR Inspect Hawk runs Inspect AI evals at scale on isolated infrastructure and is the catalog tool for repeating that as an operator test harness. An allowlist miss is a failed acceptance test, not a prompt to “try harder.”

Sub-Agents
Host Authority (Sentinel)
Skills Required
METR Inspect HawkInspect AIEgress allowlists
Human TouchpointDo not expand auto-allow until evals and allowlists pass
06

Purchase path via single-use scoped cards

Led by: Purchase Path Operator

Muse treats checkout as high-risk: stored-card sites get a human approval with exact purchase details; new merchants use a wallet. At launch Muse partnered with Stripe Link: a single-use card scoped to merchant, amount, and a limited validity window is what hits the merchant — not the everyday card. Stripe’s own agent wallet: OAuth to Link, create a spend request, human approves in Link, agent receives a one-time card or shared payment token (SPT). Same rule as the payments playbook: capability, not PAN.

Sub-Agents
Host Authority (Sentinel)Paying Agent
Skills Required
Stripe LinkSingle-use scoped cards
Human TouchpointApprove each spend request (Stripe Link today requires per-request review)