P
Permission Engine
Define which assets, contracts and actions are available to an agent. Scope is visible before a request reaches an execution path.
Access / scope / rules
CONTROL INFRASTRUCTURE / ARC
Safe execution for onchain agents.
SENTRY gives autonomous workflows an explicit operating envelope before execution begins. Define what an agent can access, how much it can move, and which actions remain inside policy.
PERMISSIONS · LIMITS · VERIFIABLE ACTIONS
Permission engineTransaction boundariesPreflight policy checksReviewable execution
01 / CONTROL SURFACE
Authority should be explicit before an agent acts.
Autonomous systems become easier to operate when access, financial limits, approved targets and execution rules are visible in one place. SENTRY turns those controls into a readable policy layer for workflows on Arc.
L
Transaction Boundaries
Apply per-request ceilings and recurring budgets so automated operations stay inside a known financial envelope.
Limits / budgets / cadence
R
Reviewable Records
Keep the request, policy decision and final outcome readable so operators can understand what happened after execution.
Decisions / logs / proof
02 / EXECUTION MODEL
A clearer path from intent to execution.
SENTRY separates an agent’s intent from the policy decision that determines whether the action is eligible to continue. The model keeps the control path inspectable without hiding it behind a black box.
01 / REQUEST
A
Agent submits a request
Target, asset, amount and intended action enter the control surface.
RECEIVED
02 / EVALUATE
P
Policy evaluates scope
Allowlists, limits and action constraints are checked against active rules.
CHECKED
03 / PREFLIGHT
S
Preflight decision
The request is evaluated before an execution layer is engaged.
PREFLIGHT
04 / EXECUTE
X
Controlled execution path
Eligible actions can move forward through the connected environment.
CONTROLLED
05 / RECORD
✓
Outcome stays reviewable
Decision context and final status remain available as an operational record.
RECORDED
03 / OPERATOR VIEW
One policy view. Less ambiguity.
A compact operating surface keeps the most important boundaries visible to the people responsible for the agent, without requiring them to parse raw calls or scattered configuration.
AGENT / TREASURY-01ACTIVE
NETWORKARC
MAX TX250 USDC
DAILY1,000 USDC
ASSETSUSDC / WETH
STATUSREADY
ACTIVE RULE SET04 RULES
Contract allowlistRestrict execution to approved targets
ONTransaction ceilingApply per-request value limits
ONDaily operating budgetManage cumulative exposure
ONPreflight simulationEvaluate before execution proceeds
ON04 / USE CASES
For workflows where limits matter.
SENTRY is designed around a simple operating principle: automation should have enough authority to work, but not more authority than it needs.
Control how capital can move.
Define approved assets, contract targets and maximum transaction values before treasury actions can proceed.
Constrain where strategies execute.
Limit venues, action types and operating ranges so automated strategies remain inside intended parameters.
Separate automation from unlimited authority.
Keep recurring operational tasks inside a controlled and inspectable execution envelope.
Make agent behaviour easier to review.
Give builders and operators a readable model for permissions, policy decisions and execution outcomes.
SENTRY / BUILT ON ARC
Build automation with boundaries.
Explore the policy model, inspect the operator surface and test the browser-based simulator behind SENTRY.
usesentry.xyz