Softlander Book a call
Agent security and cybersecurity

Cross-agent attack analysis

Identify how one agent can manipulate another agent's decisions

External or lower-trust agents participate in a shared workflow. The decision is how to identify how one agent can manipulate another agent's decisions. The work begins with your actual workflow and looks for the evidence that separates intended behaviour from a consequential exception. It is particularly relevant to multi-agent security architects and coordination owners. A useful outcome is an explicit account of what was examined, what remains uncertain, and which decision the evidence can support.

Book a 30 minute readiness callBrowse this capability group

The agent security relationship

  1. Starting evidenceAgent roles

    Establish the basis for compromised role.

  2. Decision boundaryMessage provenance

    Check the origin of delegated instructions.

  3. Decision artefactCross-agent attack map

    Retain the scoped finding and its unresolved dependencies.

Proposed working model. A bounded challenge cannot establish exhaustive protection.

Decision

Frame the decision around compromised role

Questions and the evidence they need

Criterion / evidenceCompromised role
Inspect agent roles for the relevant constraint or observation.
Criterion / evidenceMessage provenance
Inspect message provenance for the relevant constraint or observation.
Criterion / evidenceReceiving boundary
Inspect shared resources for the relevant constraint or observation.
Read each criterion against the available evidence before deciding that additional work is needed.

The decision is how to identify how one agent can manipulate another agent’s decisions. This is relevant to multi-agent security architects and coordination owners when external or lower-trust agents participate in a shared workflow. Name the affected workflow, the consequence of a wrong decision, and the owner who can act on the result. The initial discussion should establish which evidence is already available and which uncertainty is material enough to investigate.

The specific failure hypothesis is that a compromised participant sends a fabricated reviewer conclusion that the coordinator treats as approval. It remains a hypothesis until examined in the deployment. Existing internal tests or operating records may already resolve it; identify the uncovered question before commissioning additional work.

Mechanism

A peer message should not inherit the sender's claimed authority

Relationships within agent security

Source or conditionCompromised role
Define the participant assumed under attacker control.
Crossing or choiceMessage provenance
Check the origin of delegated instructions.
Consequence or decisionReceiving boundary
Limit downstream action to verified authority.
Observe the control and resulting business effect. Each relationship needs evidence in the actual scope.

What this specialty examines

Cross-agent attack analysis treats messages and specialist outputs according to their origin and authority. An agent’s statement that a review passed is not automatically an accountable approval. Identify which participants can recommend, verify, approve or execute.

The review should model attacker control by participant and test synthetic misleading messages across role boundaries. Identify the normal route, relevant exceptions and the authoritative record of the outcome. Inspect the relevant business-system record, enforcement event or process evidence alongside any reported status. Keep incomplete cases visible, with an explanation of what their absence prevents the owner from concluding.

The tradeoff to make explicit

For this decision, specialized collaboration improves coverage while role labels need enforceable meaning. Compare the available design or operating choices against the same intended outcome. If one option removes a required control or changes the task boundary, describe that difference before treating its apparent simplicity, speed or lower effort as an improvement. The decision should expose the tradeoff instead of burying it in an aggregate score.

Preparation

Connect the working inputs to the boundary being examined

Evidence packet for compromised role

Illustrative working format

BasisAgent roles
Identify the accountable role and verify its decision rights or availability.
ContextMessage provenance
Connect the deployed artefact to its inputs, approval, and verifiable revision.
ObservationShared resources
Establish permitted access, provenance, and which task conditions it represents.
DependencyDelegation rules
Locate transfer of responsibility, permitted authority, and unfinished-work ownership.
Illustrative preparation record. Availability, revision, and handling need confirmation.

Start with agent roles, message provenance, shared resources and delegation rules. Identify the owner, revision and operating context of each source. Sanitised examples may support initial design, but differences from actual conditions should remain explicit. Where direct access is unsuitable, a customer-controlled evidence or execution route may be considered, with agreed handling and retention.

The scope also needs the security decision owner, agreed expectations for a representative threat scenario, and the authority to inspect the relevant boundary. Specify what evidence would distinguish the failure in which a compromised participant sends a fabricated reviewer conclusion that the coordinator treats as approval from acceptable behaviour. If stakeholders disagree about intended behaviour, resolve the interpretation before classifying the observed difference as a defect.

How the work proceeds

Carry compromised role through to an accountable finding

From the scoped material to a decision record

  1. Establish conditionsAgent roles

    Define the hypothetical attacker’s starting influence.

  2. Examine the relationshipMessage provenance

    Apply the scoped check to receiving boundary.

  3. Retain the findingCross-agent attack map

    Record the evidence, open dependencies, and the owner’s next decision.

Proposed sequence; resolve interpretation or access gaps before relying on the next step.

Define the controlled threat scenario

Review agent roles, message provenance, shared resources and delegation rules with multi-agent security architects and coordination owners. State what the hypothetical attacker controls, the protected action and the systems included in the authorised exercise. Prepare synthetic records and harmless markers where appropriate, with stopping conditions for unexpected effects. The preparation output is a scenario brief that distinguishes an influence attempt from a prohibited action actually reaching a business resource.

Challenge the path and review effects

Test misleading role claims, fabricated evidence references and conflicting conclusions under controlled conditions. Track how the coordinator resolves disagreement and whether it checks provenance before granting additional authority. Preserve uncertainty instead of forcing consensus.

Select a control response and retest

Build a cross-agent attack map with message-authenticity and authority findings from the reviewed observations. Separate agent behaviour, application enforcement and the resulting resource state so the recommended change reaches the responsible layer. Choose a bounded retest that reproduces the original condition and checks legitimate behaviour after remediation. If the required investigation exceeds the agreed technique or expertise, retain the unresolved path and identify the qualified review needed.

Indicative artefact

Make the cross-agent attack map usable by its owner

Cross-agent attack map

Illustrative working format

Annotated fieldCross-agent attack map
Identify the assessed workflow, revision, and decision boundary.
Annotated fieldMessage-authenticity
Link supporting and contrary observations to inspectable evidence.
Annotated fieldAuthority findings
Link observations and counterevidence to the case, configuration, and conclusion limit.
Illustrative artefact structure, not a completed engagement or an observed result.

A useful decision artefact is a cross-agent attack map with message-authenticity and authority findings. Supporting material may include selected cases, configuration or process records, evidence references, a disagreement log and prioritised actions. Retain the case selection rationale and material disagreements alongside the result; otherwise the result may be mistaken for complete coverage.

Illustrative scenario

Examine receiving boundary through a constructed exception

A constructed case and its decision paths

Illustrative scenario

  1. SituationCompromised role

    A specialist sends a misleading instruction to a more privileged coordinator.

  2. Question to examineMessage provenance

    Trace the specialist message's origin and claimed delegation.

  3. Conditional next actionReceiving boundary

    Require verified authority before the coordinator follows the instruction.

Fictional example. No executed test, customer engagement, or measured outcome is implied.

Consider a fictional situation: a research agent’s output is formatted like a security review and a coordinator treats it as permission to proceed. The scenario is illustrative, with no client engagement, executed test or measured improvement implied.

For the false security review, require an approval source that the research agent cannot manufacture. The coordinator may still use the research as input, but should not promote it into permission. That distinction belongs in both contracts and enforcement.

Scope and next action

Keep the conclusion within the agreed threat scenario

Confirm fit before the next commitment

Included boundaryCompromised role
This review is limited to the agreed threat scenario.
PrerequisiteAgent roles
The relevant owner must confirm access and the intended operating conditions.
Claim limitCross-agent attack map
A bounded challenge cannot establish exhaustive protection.
Bring the triggering change and existing evidence; resolve the dependency that could alter the decision.

Authorised scenario testing is not a claim of exhaustive penetration testing or guaranteed protection. Intrusive security work requires qualified specialists and a separately agreed scope. Any engagement depends on an agreed scope, qualified team, authorised access and confirmed availability. Bring the affected workflow, its current configuration or process, the triggering change and the evidence already available. The immediate scope question is prompted by the situation in which external or lower-trust agents participate in a shared workflow; confirm which missing observation would change the owner’s next action.

Ways to engage

Related engagement scopes

Agent Release Readiness Review

The Agent Release Readiness Review is a fixed-scope, three-week independent test of one AI agent workflow that shows a release board what the agent does, what it is allowed to do and what happens when it is wrong, and ends in a signed recommendation.

3 weeksRead about this offer

AI Agent Runtime Security Review

Make every delegation explicit. Examine who the agent acts for, where it can execute and which tools it can reach when instructions alone are not enough.

Read about this offer

AI Agent Red Team

Challenge the boundaries that matter to your workflow. Test agreed attack scenarios against agent instructions, tools and data access.

Read about this offer

Continue exploring

Related insights and resources

Agent security6 MIN READ

The permission boundary behind every agent action

An agent can only be evaluated in the context of what its tools allow. Examine identity, delegated authority and action boundaries before you put a workflow in motion.

Whitepaper5 MIN READ

Before your agent can act, define its authority

A working paper for the moment an AI system moves from advice to action. Map identity, permissions, approvals and recovery into an authority model your team can review and test.

Agent security6 MIN READ

Running an AI-agent incident tabletop exercise

A useful tabletop follows one plausible agent failure through detection, containment, evidence, business recovery and accountable return to service.

A little more clarity

Questions to consider.

Start with the questions that shape a better decision.

Should another agent's recommendation be treated as an approval?

Only when a defined approval process authorises that role and verifies the decision's provenance. Recommendations, analysis and permission to act should remain distinct even when exchanged in similarly formatted messages.

Which role records define cross-agent trust?

Start with agent roles, message provenance, shared resources and delegation rules. Record their revision, owner and relationship to the intended workflow. Missing context should be identified before interpreting outcomes; a sanitised fixture is useful only when its relevant differences from the operating case are understood.

How should false peer approvals be tested?

Model attacker control by participant and test synthetic misleading messages across role boundaries. State the hypothetical attacker's control and inspect attempted actions, tool enforcement and actual effects separately. A refusal, a blocked call and a successful prohibited action are different results. Preserve unsuccessful challenges as well as adverse findings.

What should a cross-agent attack map show?

The proposed output is a cross-agent attack map with message-authenticity and authority findings. Include the selected conditions, supporting observations, unresolved differences and next decision. A reviewer should be able to follow each important conclusion back to its evidence and understand the limits without depending on an undocumented explanation from the author.

When should a specialist conclusion require verification?

Verify when a specialist conclusion would grant authority, satisfy a control gate or materially change a consequential decision. Descriptive peer output should not become approval without the defined provenance and role checks.

What if existing evidence already answers the question?

Review its scope, provenance and configuration before proposing new work. Existing work from multi-agent security architects and coordination owners may already be sufficient. In that case, capture the decision, unresolved limitations and the change that would require another review. Additional assessment should address a real gap.

What would make the findings inconclusive?

Missing agent roles, message provenance, shared resources and delegation rules can prevent a reliable interpretation, depending on the case. Record the specific missing observation and its implication. The owner can supply evidence, narrow the claim or defer the dependent decision; an incomplete case should not silently be counted as a pass.

Start with the decision you need to make.

Tell us about the agent workflow you want reviewed, and where you need greater clarity. Scope and availability are confirmed together.

Book a 30 minute readiness call Get the agent release checklist

Connected disciplines

Agent identity spoofing assessment

Verify that agent identity claims cannot substitute for authentication. The review connects peer registries, identity proof, issuer rules and role-assignment records to a bounded review of agent security, giving the security decision owner a clearer basis for the next action.

Memory poisoning tests

Examine whether persistent memory can carry attacker-controlled instructions. The review connects memory write paths, provenance, retrieval rules and correction controls to a bounded review of agent security, giving the security decision owner a clearer basis for the next action.

Retrieval poisoning tests

Test how corrupted retrieval sources influence agent outputs and actions. The review connects corpus permissions, indexing rules, ranking behaviour and source-version records to a bounded review of agent security, giving the security decision owner a clearer basis for the next action.