Skip to main content
Version: 11.5.0

Security Controls Reference

This page lists the controls HYPR AgentPass™ enforces on agent traffic, written for a security reviewer or auditor. Each row makes one claim: what the gateway does, the risk it answers, and where to read more. Each control keeps its AP- identifier, so you can cite it. Most claims link to the page that documents them. The rest name your HYPR representative as the source to confirm with.

The sections follow the questions a reviewer asks of an agent gateway:

  • What an agent may do
  • How an agent is stopped
  • How one session is contained
  • How one context is kept from another
  • How instructions and secrets are protected
  • How access is confirmed over time
  • How sensitive data is handled
  • How restrictions follow data from one call to the next
  • What happens to information an agent retrieves
  • How instructions injected into content are contained

Where a control reaches the edge of what the gateway can see, the row says so instead of claiming more.

Agent Privileges​

ControlWhat it preventsHow AgentPass does itWhere to read more
AP-PRIV-1: Authorization is decided by people, in the consoleAn agent's permissions being set anywhere other than by an administrator, or changing without a recordAllow, block and require-approval policies, and the per-role tool grants, are written by administrators in the console, which is itself governed by roles. The gateway enforces only what the console holds, and the policy list records how each rule was authoredPolicies · Roles and Permissions · Controlling MCP Tool Access by Role
AP-PRIV-2: A blocked action stops at the gatewayA denied action reaching the tool or the model anywayEvery tool call and every model request is checked against your policies before it is forwarded. A block ends the request at the gateway; nothing reaches the targetHow Policies Are Evaluated · Governing Inference Traffic
AP-PRIV-3: Tool access is closed by defaultAn agent calling, or even discovering, a tool its role was never grantedA tool not granted to the supervisor's role is neither listed to the agent nor callable. The list and the call check read the same decision, so they cannot disagree, and a newly discovered tool starts with no accessControlling MCP Tool Access by Role
AP-PRIV-4: An agent cannot administer the gatewayAn agent using its own credential to change policies, roles, or its own permissionsAn agent's credential opens only the connector endpoints it was given. The console accepts only a signed-in person's identity, and a workload agent can never be issued the administrator roleSigning In to the Console · Creating a Workload Agent
AP-PRIV-5: A workload agent's reach is its own, named boundaryAn unattended agent inheriting the access of the person who set it upA workload agent's own role, not its owner's, decides its tools, and the connectors named on its credential, matched exactly, decide where that credential is accepted. Anything else is refusedWorkload Agents · Creating a Workload Agent
AP-PRIV-6: Revocation is immediate, everywhereA disabled agent or a withdrawn grant continuing to work while a change propagatesEvery check reads the agent's status and grants afresh on each request, across the whole deployment. The next request after a change sees the changeManaging Workload Agent Credentials · Roles and Permissions · Turning Off an Agent
AP-PRIV-7: A privileged action needs a person, on their own deviceA sensitive action proceeding without a verifiable human decisionA require-approval policy holds the call until a supervisor approves it, on their phone or in the console. With Require signed approvals set to Required (production), every approval must be signed with that supervisor's own passkey, so the record shows the decision was theirs. Under Optional (soak), which a new organization starts on, an approval can be recorded without a signature. A denial needs no passkey under either setting Approving and Denying Agent Actions · Setting Up Your Approval Passkey
AP-PRIV-8: A shared agent acts as the person using itOne person's tool call running with another person's accessWith per-user isolation on, each call an AWS Bedrock AgentCore agent makes runs as the person in that session, with their connected accounts, consents and approvals, and only for that session. A call whose person cannot be determined is refused rather than run as anyone elseAWS Bedrock AgentCore Agents

Stopping an Agent​

ControlWhat it preventsHow AgentPass does itWhere to read more
AP-KILL-1: One action stops any agentNeeding a different procedure for each kind of agentTurn off, on the Agents page, works the same way for every agent the page lists: local agents, cloud agents, AWS Bedrock AgentCore agents and workload agents. It is in each row's ⋮ menu and in the agent's Details drawerTurning Off an Agent
AP-KILL-2: Disabling takes effect on the very next requestA disabled agent finishing "one more" callA disabled agent's next call is refused at the gateway, whatever credential it presents. On an AWS Bedrock connector, that includes Observe only and Require Approval modes. In Disabled mode no caller is recorded, so there is no agent to refuse Turning Off an Agent · Managing Workload Agent Credentials
AP-KILL-3: It works across the whole deployment at onceA stop that reaches one part of the deployment before anotherAgent status is read afresh on every request, across the whole deployment, so there is no propagation to wait out and nothing to restartTurning Off an Agent
AP-KILL-4: Running work is interruptedA disabled agent completing a session already in progressTurning off an Anthropic cloud agent interrupts and archives its active sessions, so they cannot be resumed. Turning off an AWS Bedrock AgentCore agent stops its runtime sessions in AWS Anthropic Cloud Agents · AWS Bedrock AgentCore Agents
AP-KILL-5: Schedules do not fire for a stopped agentAn automation quietly running an agent that has been disabled or removedA scheduled or triggered run against a disabled, archived or removed agent is recorded as skipped and never invokes the agent. A run in progress when the agent is turned off ends Automating Cloud Agents
AP-KILL-6: Offboarding a person revokes everything they heldA departed person's agents, sessions and connected accounts continuing to actWhen an administrator offboards a person from the console, or your directory deprovisions them, every managed-agent credential issued to them, their running sessions, their connected-account tokens, and their group memberships and agent assignments are revoked at onceUsers
AP-KILL-7: Only an authorized person can flip the switchAn agent being turned off, or back on, by someone outside its scopeTurning an agent off or on requires a console permission. At All scope it reaches every agent. At Own scope it turns off only the agents the person owns, and turns one back on only if that person turned it off. The rule is the same in the console and the mobile app Roles and Permissions · Who Can Turn an Agent Off and On · Approving and Denying Agent Actions
AP-KILL-8: Every flip is attributableNot knowing who stopped an agentAgentPass records who turned an agent off, when, and the reason they gave, and shows them in the agent's Details drawer and History. Turning mobile access on or off, and offboarding a person, record who did it Turning Off an Agent · Agent Details
AP-KILL-9: A whole organization can be containedAn incident that outgrows a single agentA platform administrator can disable an entire organization, after which every request to that organization's gateway address is refusedConfirm with your HYPR representative
AP-KILL-10: Detection has somewhere to go, and can actSuspicious behavior being noticed with no way to act on itSession Watch evaluators review sessions and report observations, and every policy block is recorded on the connector's Policies tab. Detection is observational by default, and an administrator or supervisor acts on it. Where you set a session investigator to enforce, its verdicts can also contain a session automatically, failing open by default. See Session-Level ContainmentSession Watch · The Session Investigator · Approving and Denying Agent Actions

Session-Level Containment​

Stopping an Agent acts on an agent identity. These controls act on a single live session instead. A reviewer, or a session investigator you set to enforce, can pause, resume or terminate one coding-agent session without taking the agent away from everyone else.

ControlWhat it preventsHow AgentPass does itWhere to read more
AP-SESSION-1: One session can be frozen without stopping the agentHaving to disable a whole agent to contain one person's runA paused session has every tool call refused at the gateway before it reaches the tool, reads included, while the agent keeps running and can explain itself. Other sessions of the same agent are untouchedPausing and Stopping a Session
AP-SESSION-2: Terminate is finalA terminated session being quietly re-openedA terminated session cannot be paused, resumed, or re-opened; the person starts a new session to continuePausing and Stopping a Session
AP-SESSION-3: The freeze holds everywhere at onceA pause reaching one part of the deployment before anotherThe control state is read afresh on every request, across the whole deployment, so a paused session is frozen everywhere at once, with nothing to wait outPausing and Stopping a Session
AP-SESSION-4: A call already in flight is accounted forA pause silently dropping a call that was already runningThe freeze acts on the next call; a call already past the gate runs to completion and is recorded, and everything after the pause is frozenA Call Already in Flight Finishes
AP-SESSION-5: Automatic containment is opt-in and measuredAn over-eager investigator freezing work it should notThe investigator runs off the tool-call path and, by default, only observes, recording what it would do. It acts only in enforce mode, at or above a confidence threshold you set, and never freezes a session that is not actually liveThe Session Investigator
AP-SESSION-6: Uncertainty fails open by defaultA flaky detector becoming a fleet-wide work stoppageWhen the investigator is uncertain or unavailable, the session continues and the event is logged. Failing closed, which pauses the session instead, is an explicit opt-inFail-Open and Fail-Closed
AP-SESSION-7: Nobody lifts their own pauseThe identity a pause was meant to contain undoing itThe session's own identity cannot resume it; a distinct reviewer must, and that reviewer must also hold the right permission. Both checks hold on every path: the console, the mobile app and the APIEscalation and Signed Resume · Approving and Denying Agent Actions
AP-SESSION-8: A pause does not hang foreverA frozen session left unresolved because no one answeredEach pause carries a time limit; if it passes with no response, the session resolves by the fail mode, resuming when failing open and terminating when failing closed, and the outcome is recordedIf No One Responds
AP-SESSION-9: A hostile verdict contains the identity, not just the sessionA compromised identity opening a fresh session to bypass the freezeA terminating verdict can also disable the agent's identity, so the next request is refused whatever credential it presents, until a person re-enables it. It is additive, and never undoes the terminate it is added toA Hostile Verdict Disables the Identity
AP-SESSION-10: Every containment is attributable and auditableNot knowing who froze a session, or whyEvery pause, resume, and terminate is recorded against the reviewer or investigator that performed it, with its reason and time, and emitted to the audit feed; the reason also reaches the developer running the agentWho Can Pause a Session, and the Record It Leaves

Keeping Contexts Apart​

ControlWhat it preventsHow AgentPass does itWhere to read more
AP-ISO-1: Every record belongs to exactly one organizationData from one organization appearing in anotherEverything AgentPass stores is scoped to the organization it belongs to, with no default organization to fall back to. A request to a gateway address that maps to no organization is refusedData Management
AP-ISO-2: Background work stays in its organizationScheduled or deferred work losing track of which organization it servesWork the gateway performs outside a request, such as applying retention or running an automation, carries its organization with it and cannot run without oneConfirm with your HYPR representative
AP-ISO-3: Policies are per organizationOne organization's rules acting on another's trafficPolicy evaluation is keyed by organization. Only your policies run on your trafficConfirm with your HYPR representative
AP-ISO-4: Connected accounts are personalOne person's connected account being used for another person's callThe tokens the gateway holds for a connector are stored per organization, per connector, per person, and never served across any of the three. Each person connects their own accounts, and an agent acts with that person's accessConnecting MCP Servers · Anthropic Cloud Agents
AP-ISO-5: Nobody spends another person's account without an opt-inA workload agent quietly acting as its ownerA workload agent does not use its owner's connected accounts unless an administrator names that connector on the credential. Without the opt-in, such a call is refusedWorkload Agents · Creating a Workload Agent
AP-ISO-6: An isolated call runs as its own person, or not at allA shared cloud agent running a call under the wrong person's identityOn a per-user AWS Bedrock AgentCore agent, a session that is unknown, has ended, or belongs to another agent or another organization is refused, and the tool is never calledAWS Bedrock AgentCore Agents
AP-ISO-7: A credential is good only where it was issuedAn agent credential from one organization being accepted by anotherA credential issued for one organization is refused at another organization's gateway addressConfirm with your HYPR representative
AP-ISO-8: A cloud agent's credentials are bounded to its declared toolsAn agent receiving a credential for a connector it was never givenEach person and agent pair holds its own credential store, and the agent's connector list bounds what that store may containManaged Cloud Agents · AWS Bedrock AgentCore Agents
AP-ISO-9: Every call has exactly one agent and one personPolicy or audit acting on an ambiguous identityThe gateway resolves each request to one agent and one person, preferring the person running the session, before any policy runs, and that is all a policy sees. A gateway session credential is never forwarded to the targetViewing the Request Log · Attribute Reference
AP-ISO-10: Caches cannot mix organizationsCached data serving the wrong organizationCached data is kept separate per organizationConfirm with your HYPR representative
AP-ISO-11: Captured samples never hold credentialsA secret surviving in the request recordRequest and response samples are redacted before they are stored, and the raw view in the Request Log shows secrets redactedViewing the Request Log · Data Management
AP-ISO-12: Persistent agent memory is governed at the configuration levelUnreviewed changes to what a cloud agent carries between sessionsAgentPass governs who may view or change an agent's configuration, memory settings included; the contents of that memory live with your provider and never pass through the gatewayAWS Bedrock AgentCore Agents · Managed Cloud Agents

Protecting Instructions​

ControlWhat it preventsHow AgentPass does itWhere to read more
AP-PROMPT-1: Reading an agent's instructions needs a grant; changing them needs a stronger oneA person viewing or rewriting an agent's prompt beyond their roleViewing a cloud agent's configuration requires a view permission and changing it a manage permission. A permission held at Own scope reaches only the person's own agentsManaged Cloud Agents · Roles and Permissions
AP-PROMPT-2: No agent-facing surface exposes configurationAn agent reading prompts, policies or connector settings through the gatewayThe endpoints an agent can reach carry tools only. Prompts, policies and connector configuration are console-only, and no agent credential is a console identitySigning In to the Console · Connecting a Workload Agent
AP-PROMPT-3: Tool routing is invisibleAn agent learning which tools exist beyond its grantThe tool list handed to an agent is trimmed to what its role may call. Where a call is routed is decided at the gateway and never described to the agentControlling MCP Tool Access by Role
AP-PROMPT-4: A refused call receives a verdict, never the ruleA blocked agent learning how a rule is built and working around itA blocked agent receives the policy's name and the message you wrote; the condition itself never leaves the gateway. A tool refused by role is told only which tool was refusedHow Policies Are Evaluated · Controlling MCP Tool Access by Role
AP-PROMPT-5: Secrets are write-only and encrypted at restA connector secret being read back out of the consoleConnector secrets and provider keys are stored encrypted and never shown again after saving. Changing one means replacing itAnthropic Cloud Agents · Signing In to the Console
AP-PROMPT-6: Secrets are attached per call, never shown to the agentAn agent, or the tool behind it, holding a credential it should never seeThe gateway attaches the downstream credential at the moment it forwards a call. The agent never holds a credential of its own, and a gateway session credential is never passed to the targetConnecting MCP Servers
AP-PROMPT-7: Execution records are console-onlyAn agent reading audit rows, policy verdicts or approval stateRequest records, policy verdicts and approvals are read through the console under role scope, and samples are redacted before they are storedViewing the Request Log · Roles and Permissions · Data Management
AP-PROMPT-8: Guidance AgentPass adds does not advertise privileged toolsThe gateway's own instructions steering an agent toward a privileged actionThe guidance AgentPass adds to a cloud agent's instructions explains how to behave when a call is held for approval. It does not list or promote privileged toolsAnthropic Cloud Agents · Managed Cloud Agents

Agent Access Reviews​

ControlWhat it preventsHow AgentPass does itWhere to read more
AP-REVIEW-1: Every agent has a recorded lifecycle stageA retired agent quietly coming back into use, or a stage change nobody can explainEach agent is in Production, Retain data only, or Retired. Leaving Production switches the agent off, and only a Production agent can be switched on again. Every change records who made it, when, and whyLifecycle Stages
AP-REVIEW-2: Reviews recur without anyone remembering to start themAccess going unreviewed because nobody opened a reviewA schedule per stage opens a review on a set interval with no one acting. A new review opens only when the previous one for that stage is no longer open, so the same period is never reviewed twiceSet a Schedule
AP-REVIEW-3: A review covers every agent in scope, onceAn agent slipping through a review, or being decided twiceEvery agent at the review's stages is included exactly once, and a person's desktop AI apps are one item. Retired agents are never includedWhich Agents Are Reviewed
AP-REVIEW-4: Nobody certifies their own agentAn owner approving their own agent's access, or a departed person deciding anythingAn agent is never assigned to its owner. Only an active person who holds the review permission can review, and this is checked again at the moment of each decisionWho Can Do What
AP-REVIEW-5: Decisions are made on the evidenceA reviewer approving access they cannot seeWhen a review opens, AgentPass takes a snapshot of each agent's reach per connector with its most sensitive data tier, who can use it, its credential facts, and its usage over 90 days. The reviewer decides on that snapshotRead the Evidence
AP-REVIEW-6: Each decision is recorded once, by the assigned reviewerA decision that cannot be attributed, or a removal quietly reversedOnly the assigned reviewer can decide an agent, and each decision is recorded once with their name. A removal needs a reason, and a later Keep cannot undo itDecide
AP-REVIEW-7: A removal takes effect, or stays visible until it doesA reviewer's removal being recorded but never carried outA removal is applied through the same controls an administrator uses. One that cannot be applied stays listed for the review's managers, with the reason, until it appliesRemovals That Did Not Apply
AP-REVIEW-8: People are told what they need to act onA review stalling unnoticed, or an owner not knowing why access disappearedReviewers are reminded before the due date and after it until they finish. Managers are told about agents nobody can review and removals that did not apply. An agent's owner is told what was removed, by whom, and whyNotifications · What Managers Are Told
AP-REVIEW-9: An undecided agent is never counted as approvedA review closing with gaps that read as approvalsClosing a review records every undecided agent as not reviewedClose a Review
AP-REVIEW-10: Results leave in a form auditors and governance tools can useReview results that cannot be taken to an identity-governance tool, or an export that runs a formulaEach review exports as CSV or as JSON with the evidence, in fields that are added to but never renamed. Spreadsheet cells cannot start a formula, and every export is recorded. The person who runs the review program, holding Access reviews: Manage, exports the results and passes them to the auditorExport the Results
AP-REVIEW-11: Running the program and reviewing are separate permissionsA reviewer changing schedules, or a manager deciding agents assigned to someone elseSetting schedules, opening and closing reviews, and exporting need the manage permission. The review permission reaches only the agents assigned to the person holding itWho Can Do What
AP-REVIEW-12: A reviewer is a named person, with one exceptionA decision that cannot be tied to one individualEvery reviewer is a named person, except the recovery administrator while local sign-in is available: it is a shared break-glass account, so its decisions are recorded under that account and not tied to one person. This is where the control stops; to keep every decision tied to one person, name individual people as reviewersThe Recovery Administrator as a Reviewer

Handling Sensitive Data​

ControlWhat it preventsHow AgentPass does itWhere to read more
AP-DATA-1: Only approved tools runAn agent using a capability nobody reviewedA tool discovered on an MCP server starts with no access; an administrator enables it and a role grants it. On a REST API connector every operation is available once the API is imported, so there the approval is importing the connectorControlling MCP Tool Access by Role
AP-DATA-2: A tool's data carries a classification policies can readRules that can only name tools, not the sensitivity of what they handleTiers and tags are assigned to tools, by name rule, or as a connector default, and reach every policy on every call to that toolData Classification
AP-DATA-3: A call is classified by what it carriesA sensitive payload passing through a tool nobody labeledOptional detectors (card numbers, credentials, government identifiers) and your own dictionaries read a call's arguments and replies, and raise its classification Data Classification
AP-DATA-4: Access above a tier can need a person's approvalSensitive data reaching an agent without a human decisionA require-approval policy on the classification holds the call, and the approval the supervisor signs names the classificationData Classification · Approving and Denying Agent Actions
AP-DATA-5: Access above a tier can be refusedAn agent reaching data above what it should ever seeA block policy on the classification refuses the call before it is forwardedData Classification
AP-DATA-6: Unclassified traffic is not treated as sensitive, or as safeA rule for sensitive data blocking everything, or being skipped, because a call carries no labelA call nothing classified carries no classification at all, so a rule written for classified data does not match itData Classification
AP-DATA-7: A classification cannot be lowered silentlyProtection being reduced without anyone noticingThe highest matching label always wins; lowering one needs a written reason and is recorded; and a tier or tag cannot be removed from the vocabulary while anything still uses itData Classification
AP-DATA-8: Renaming a tier does not break rulesA rule quietly ceasing to match after the vocabulary is tidied upRules compare a tier's rank, and a tier's display name can change without touching its keyData Classification
AP-DATA-9: Only authorized people change classificationsSomeone who manages connectors also being able to declassify their dataReading and changing labels and vocabulary are their own console permissions, separate from managing connectorsData Classification
AP-DATA-10: An agent cannot change its own privilegesAn agent granting itself access to more dataSee AP-PRIV-4Signing In to the Console

Carrying Restrictions Across Calls​

ControlWhat it preventsHow AgentPass does itWhere to read more
AP-FLOW-1: Every hop is checkedContext moving between agents and tools without passing a boundaryEvery call from an agent to a tool passes through the gateway, where organization and per-person boundaries are checkedSee AP-ISO-1, AP-ISO-6, AP-ISO-9
AP-FLOW-2: A call's classification is recorded with itNot being able to show what sensitivity of data a call movedThe request log records each call's classification, including one raised by what came backViewing the Request Log
AP-FLOW-3: A destination is cleared only up to a levelData reaching a system that should never receive itA connector declares the most sensitive data it may receive. The ceiling alone refuses nothing: a flow policy compares each call with it, and holds or refuses a call above it before it is forwarded Data Classification
AP-FLOW-4: Adopting a flow rule changes nothing until you set a ceilingA new rule disrupting traffic before you have decided where data may goA connector with no ceiling is never matched by a flow ruleData Classification
AP-FLOW-5: What a session read follows it to its later callsAn agent reading sensitive data and sending a reworded version somewhere elseWith session tracking on, the most sensitive data a session has read is carried to its later calls on any connector, so a send is judged on what the session read, not on its wordingSession Tracking
AP-FLOW-6: Only data the agent actually received countsA refused or failed call marking a session as sensitiveA call that was refused, held, or failed upstream leaves the session as it wasSession Tracking
AP-FLOW-7: One session never inherits another's historyOne person's reading affecting another person's actionsSessions are kept apart per organization. Within one, a session is an AgentCore conversation, a cloud-agent session, an agent credential or a person. A person's desktop conversations share one record, and so do the sessions of one agent credential, kept separate per user on a shared AgentCore agent Session Tracking
AP-FLOW-8: A caller cannot reset its own sessionAn agent escaping its history by claiming a fresh sessionOnly identities the gateway verified decide the session; a cloud-agent session identifier counts only under that cloud agent's own credentialSession Tracking
AP-FLOW-9: Out-of-scope context can be refused on the way backA tool returning data the agent should not holdA post-response policy reads the classification of what came back and can withhold the replyPolicy Attribute Reference
AP-FLOW-10: An approver sees why a call was heldA supervisor approving blindA held call's approval says what the session had already readSession Tracking
AP-FLOW-11: Every boundary crossing is recordedData leaving above a destination's clearance with no record, especially when nothing stopped itEach crossing is recorded as an event with its outcome (allowed, denied or held) and counted in a daily report you can exportSession Tracking
AP-FLOW-12: Session rules step aside when the session record is unavailableSession tracking stopping all trafficSession rules depend on the session record. When it cannot be read in time, they do not apply, and other policies still runSession Tracking
AP-FLOW-13: Desktop sessions are per personOne person's desktop work escaping its history by spreading across toolsRequests from desktop tools such as Claude Code join one session per person, covering everything that person does within the idle window. The recommended rule therefore holds for approval rather than refusingSession Tracking

Handling Retrieved Information​

ControlWhat it preventsHow AgentPass does itWhere to read more
AP-RETRIEVE-1: Every retrieval is attributedNot knowing which agent, acting for whom, retrieved whatEach call is recorded with the agent, the person, the connector, the tool and a summary of its argumentsViewing the Request Log
AP-RETRIEVE-2: Combining sources cannot weaken a restrictionRestricted information being diluted into something that looks less sensitiveWithin a session, the level only rises: whatever else a later call mixes in, it is judged at the most sensitive level the session readSession Tracking
AP-RETRIEVE-3: Lineage is per session, not per factTreating a session's history as proof of where each word came fromThe gateway knows that a session read restricted data, not which words in a later call came from it, and it cannot see information carried in a model's or provider's memory between sessionsSession Tracking
AP-RETRIEVE-4: Secrets are caught before they leave in a queryCredentials or regulated data being sent to an outside service inside a search or requestDetectors and your dictionaries read a call's arguments before it is forwarded, and a policy can refuse itData Classification
AP-RETRIEVE-5: A partly read payload is never reported as cleanA large payload being treated as safe because only its beginning was checkedWhen a payload is larger than the scan window, the result says so, and a policy can act on itData Classification
AP-RETRIEVE-6: The optional classifier never receives credentialsSecrets being sent to a third-party classification serviceCredentials are removed before content is sent, and a value containing one withholds the call to the classifier entirelyData Classification
AP-RETRIEVE-7: Forwarded searches carry no agent credentialsAn outside service receiving an agent's credentials with a searchAgentPass removes the agent's credentials from forwarded calls. The destination sees your gateway as the caller, and the client's own headers pass through. For searches that must not be attributable to your organization, place an anonymizing egress in front of that connectorConfirm with your HYPR representative

Containing Injected Instructions​

ControlWhat it preventsHow AgentPass does itWhere to read more
AP-INJECT-1: A session that reads outside content is markedContent written by an outsider steering an agent unnoticedEvery connector is Untrusted until an administrator marks it Trusted, with a recorded reason. Reading from an untrusted connector marks the session untrusted for the rest of its lifeThe Rule-of-Two Breaker
AP-INJECT-2: A sensitive action after an untrusted read can need a personAn injected instruction sending data out or changing somethingThe guardrail starter policies hold an external or state-changing action, or code execution, once a session has read untrusted content. Each organization adopts them, and each starts in observe modeThe Rule-of-Two Breaker
AP-INJECT-3: The approval shows facts, not the agent's wordsA supervisor approving on the strength of text an attacker wroteThe held request's Why this was held evidence is composed by the gateway: the action and its destination, the kind of sink, the untrusted source and the sensitive data readApprovals in the Console
AP-INJECT-4: A tool that changes after approval is flaggedA reviewed tool's definition being replaced with a malicious oneA changed tool definition raises an alert for review. An administrator can also withhold changed tools from every role until the change is approvedTool-Definition Integrity
AP-INJECT-5: The breaker judges behavior, not wordingAn attack evading detection by rewordingThe breaker reasons about what a session has done, not whether a message reads like an attack. It does not cover harm carried in text alone, or what a provider keeps in its own memoryThe Rule-of-Two Breaker