Softlander Book a call
Agent security and cybersecurity

Indirect prompt-injection assessment

Examine instructions hidden inside information an agent consumes

The agent reads web pages, documents, emails or tool outputs from other parties. The decision is how to examine instructions hidden inside information an agent consumes. 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 retrieval and application security teams. 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 evidenceContent provenance

    Establish the basis for source provenance.

  2. Decision boundaryIngestion boundary

    Track its influence on context and planning.

  3. Decision artefactIndirect injection assessment

    Retain the scoped finding and its unresolved dependencies.

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

Decision

Frame the decision around source provenance

Questions and the evidence they need

Criterion / evidenceSource provenance
Inspect content provenance for the relevant constraint or observation.
Criterion / evidenceIngestion boundary
Inspect retrieval paths for the relevant constraint or observation.
Criterion / evidenceAction control
Inspect trust labels 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 examine instructions hidden inside information an agent consumes. This is relevant to retrieval and application security teams when the agent reads web pages, documents, emails or tool outputs from other parties. 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 retrieved text is treated as authority to change the task or disclose information. 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

Retrieved content must not become execution authority

Relationships within agent security

Source or conditionSource provenance
Retain where the outside instruction originated.
Crossing or choiceIngestion boundary
Track its influence on context and planning.
Consequence or decisionAction control
Inspect the operation it attempts to redirect.
Observe the control and resulting business effect. Each relationship needs evidence in the actual scope.

What this specialty examines

Indirect injection assessment follows untrusted content through retrieval and interpretation. Identify where documents, messages or tool responses acquire apparent authority. Content provenance should remain distinguishable from the authenticated instruction that defines the task.

The review should place benign adversarial markers in controlled source content and trace how they influence planning and tool use. 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, rich external information improves task context while its instructions must remain distinguishable from authorised requests. 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 source provenance

Illustrative working format

BasisContent provenance
Connect the deployed artefact to its inputs, approval, and verifiable revision.
ContextRetrieval paths
Check its relationship to the relevant workflow and expose missing context.
ObservationTrust labels
Name the custodian and the conditions requiring a refreshed version.
DependencyConsequential action rules
Confirm the applicable revision and who resolves disputed interpretation.
Illustrative preparation record. Availability, revision, and handling need confirmation.

Start with content provenance, retrieval paths, trust labels and consequential action 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 retrieved text is treated as authority to change the task or disclose information 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 source provenance through to an accountable finding

From the scoped material to a decision record

  1. Establish conditionsContent provenance

    Define the hypothetical attacker’s starting influence.

  2. Examine the relationshipIngestion boundary

    Apply the scoped check to action control.

  3. Retain the findingIndirect injection assessment

    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 content provenance, retrieval paths, trust labels and consequential action rules with retrieval and application security teams. 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

Use controlled documents containing harmless redirection markers. Observe selection, summarization, planning and tool parameters. Test whether quoting the hostile instruction differs from following it, and whether a downstream control prevents the requested action.

Select a control response and retest

Build an indirect injection assessment tied to ingestion and action boundaries 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 indirect injection assessment usable by its owner

Indirect injection assessment

Illustrative working format

Annotated fieldIndirect injection assessment
Identify the assessed workflow, revision, and decision boundary.
Annotated fieldIngestion
Link supporting and contrary observations to inspectable evidence.
Annotated fieldAction boundaries
Name the missing evidence, responsible owner, and condition for revisiting the decision.
Illustrative artefact structure, not a completed engagement or an observed result.

A useful decision artefact is an indirect injection assessment tied to ingestion and action boundaries. 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 action control through a constructed exception

A constructed case and its decision paths

Illustrative scenario

  1. SituationSource provenance

    A retrieved page contains an instruction aimed at a connected business tool.

  2. Question to examineIngestion boundary

    Preserve source provenance while following the planted instruction's use.

  3. Conditional next actionAction control

    Treat untrusted content gaining action authority as the relevant failure path.

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

Consider a fictional situation: a supplier brochure includes an instruction to send an internal summary to an unrelated address. The scenario is illustrative, with no client engagement, executed test or measured improvement implied.

For the supplier brochure, the legitimate task might be extracting product details. An embedded request to email internal material should remain document content. The test should inspect whether any outbound action was proposed or executed, not merely whether the text was retrieved.

Scope and next action

Keep the conclusion within the agreed threat scenario

Confirm fit before the next commitment

Included boundarySource provenance
This review is limited to the agreed threat scenario.
PrerequisiteContent provenance
The relevant owner must confirm access and the intended operating conditions.
Claim limitIndirect injection assessment
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 the agent reads web pages, documents, emails or tool outputs from other parties; 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.

Why test retrieved content separately from user prompts?

Retrieved content has a different trust origin and may be encountered during a legitimate task. Test whether the agent treats source text as evidence or as authority to alter instructions, actions or recipients.

Which sources can carry indirect instructions?

Start with content provenance, retrieval paths, trust labels and consequential action 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 retrieved redirection be traced into tools?

Place benign adversarial markers in controlled source content and trace how they influence planning and tool use. 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 belongs in an indirect-injection assessment?

The proposed output is an indirect injection assessment tied to ingestion and action boundaries. 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 source content be treated as untrusted?

Treat source content as untrusted instructions when it originates outside the authority defining the task. It may remain useful evidence, but should not independently change permissions, recipients or workflow goals.

What if existing evidence already answers the question?

Review its scope, provenance and configuration before proposing new work. Existing work from retrieval and application security teams 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 content provenance, retrieval paths, trust labels and consequential action 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

Tool and capability abuse testing

Understand whether legitimate tools can be combined for unintended actions. The review connects tool inventories, permitted use cases, action sequences and resource constraints to a bounded review of agent security, giving the security decision owner a clearer basis for the next action.

Agent privilege-escalation testing

Test whether agent authority can increase across a workflow. The review connects permission graphs, delegation paths, privileged interfaces and test identities to a bounded review of agent security, giving the security decision owner a clearer basis for the next action.

Cross-agent attack analysis

Identify how one agent can manipulate another agent's decisions. The review connects agent roles, message provenance, shared resources and delegation rules to a bounded review of agent security, giving the security decision owner a clearer basis for the next action.