LEN-0120
Boundary Compliance Lens
Evaluates observed structure against declared boundary posture, including allow, block, and exception rules.
Primary Pattern Matches
Overreach
A structural condition where actions, effects, authority, or modifications extend beyond declared scope without authorization.
Lens Application
The Boundary Compliance Lens is a primary inspection mechanism for Overreach because it focuses on whether observed behavior stays within declared allow, block, and exception boundaries. It helps surface scope expansion, unauthorized effects, or modifications that exceed stated boundary posture.
Inspect For
- Actions operating outside declared scope
- Effects crossing blocked or restricted boundaries
- Exceptions used without clear authorization
- Modifications extending beyond approved limits
Avoid
Treat boundary mismatch as an inspection signal requiring review, not as proof that Overreach has occurred.
Secondary Pattern Matches
Boundary Leakage
A structural condition where effects, permissions, state, data, or authority cross a declared boundary without an authorized exception.
Lens Application
The Boundary Compliance Lens applies indirectly to Boundary Leakage by checking whether effects, permissions, state, data, or authority cross declared boundaries in ways that match the stated boundary posture. It helps surface crossings that lack allow rules, appear to bypass block rules, or require exception review.
Inspect For
- Cross-boundary flows without matching allow rules
- Permissions or authority extending beyond declared scope
- State or data movement relying on unclear exceptions
- Block rules bypassed by informal paths
Avoid
Treat boundary posture mismatches as review signals, not proof that Boundary Leakage has occurred.
Compatibility Violation
A structural condition where a graph, object, behavior, or runtime state exceeds a declared compatibility envelope without an authorized exception.
Lens Application
The Boundary Compliance Lens applies indirectly to Compatibility Violation by helping inspect whether observed behavior, graph shape, object state, or runtime posture sits outside declared allow, block, or exception boundaries. It can surface compatibility concerns when boundary rules and compatibility envelopes overlap, but should not be treated as a standalone determination.
Inspect For
- Declared allow or block rules near compatibility limits
- Runtime states outside expected boundary posture
- Missing, expired, or unauthorized exceptions
- Boundary rule changes that expose compatibility drift
Avoid
Treat boundary noncompliance as an inspection signal, not sufficient evidence that Compatibility Violation is present.
Threshold Breach
A structural condition where an observed metric, value, state, or condition exceeds a declared quantitative or qualitative threshold.
Lens Application
As a secondary lens, the Boundary Compliance Lens supports inspection of Threshold Breach by checking whether a reported limit-crossing condition aligns with stated boundary posture. It can surface mismatches between allowed ranges, blocked states, exception handling, and observed metric behavior without treating the breach itself as fully established.
Inspect For
- Declared allow, block, or exception rules
- Observed values near or beyond stated limits
- Boundary conditions that change threshold interpretation
- Exception paths masking repeated breaches
Avoid
Treat boundary noncompliance as an inspection signal, not as confirmation that Threshold Breach has been established.
Related Issues
Action Changed Something Else Too
An AI or agent action makes the requested change but also changes another object, field, file, state, rule, or workflow element that was not supposed to change.
Action Triggered by Confidence Score
A confidence score, certainty label, risk level, or probability-like value triggers an action without enough approval, calibration, or authority control.
Agent Keeps Expanding the Task
The agent repeatedly expands the task, plan, scope, or next-step list instead of completing the declared work.
Agent Modified Unrelated State
An agent changes state outside the requested task, target, file, record, workflow, or authorized scope.
Agent Permission Expands Over Steps
An agent begins with limited permission but gains, assumes, or exercises broader authority as the workflow continues.
AI Adds Work Not Requested
The AI adds tasks, steps, analysis, checks, changes, or follow-up work that the user did not ask for.
AI Memory Has No Governance
Saved or persistent AI memory affects output without clear rules for ownership, scope, review, update, expiry, or removal.
AI Memory Updated Without Asking
AI memory, saved context, preference, or durable state is updated without the user clearly asking for or approving that update.
AI Output Breaks Parser
The AI output causes a parser, validator, importer, or structured-output consumer to fail.
AI Touches Unrelated Scope
The AI affects, edits, analyzes, changes, or reasons over material outside the scope of the requested task.
AI Uses External Information When Forbidden
The AI uses outside knowledge, sources, tools, memory, or assumptions after the task forbids external information or limits the allowed source set.
Automation Skips Required Approval
An automated AI or workflow step proceeds past an approval gate that should have been required before action.
Behavior Does Not Match Declared Role
The AI or agent behaves outside, below, or differently from the role, authority, responsibility, or permission posture declared for it.
Citation Points to Wrong Source
A citation, reference, link, or source pointer is present, but it points to the wrong source, wrong passage, wrong document, or unsupported evidence.
Context Leaks Between Tasks
Context, assumptions, constraints, examples, files, or decisions from one task affect another task where they should not apply.
Declared Owner Cannot Control Outcome
A person, role, system, or policy is declared responsible for an outcome but does not have the actual authority or control needed to govern it.
File-Bounded Task Uses Outside Content
The AI is asked to work only from a specific file or document but uses content, assumptions, or sources outside that file.
Hallucinated Fields
The AI adds fields, keys, attributes, columns, or structured elements that were not declared, requested, or allowed by the expected schema.
Hidden Rule Overrides Visible Instruction
A hidden, upstream, system, policy, tool, or product rule changes or overrides the visible instruction the user expects the AI to follow.
Local Exception Grows Into Policy
A local exception, special case, or one-off allowance begins to function like a general policy.
Local Rule Spreads to Broader Cases
A rule intended for one local case, file, context, user, workflow, or exception begins affecting broader cases.
Model and Workflow Disagree on Next Step
The AI model recommends or selects a next step that conflicts with the workflow state, required handoff, routing rule, or process sequence.
Output Breaks After Model Change
Output that previously worked begins failing after a model, mode, runtime, or product behavior changes.
Output Breaks the Next Step
The AI output looks acceptable by itself but cannot be used by the next tool, workflow step, parser, reviewer, or downstream consumer.
Output Changed Without Declared Change
Output shape, content, format, fields, or behavior changes without a declared change to the prompt, schema, model, workflow, or governing rule.
Output Exceeds Length Limit
The AI output exceeds a declared length, token, word, character, section, field, or size limit.
Policy Exception Spreads Too Far
A narrow policy exception, allowance, or special case spreads beyond its intended scope and begins governing broader cases.
Prompt Changed but Workflow Did Not
A prompt changes but the workflow, parser, review step, routing rule, or downstream expectation still assumes the old prompt behavior.
Prompt Does Not Say What to Exclude
The prompt declares what to include but does not declare what should be excluded, allowing unwanted scope, sources, content, or actions into the result.
Retrieval Exceeds Evidence Limit
The AI retrieves, uses, cites, or considers more evidence than the task permits or more than the review surface can support.
Review Outcome Changes Unrelated Environment
A review result, approval, rejection, or classification changes state outside the environment, case, file, or workflow it was meant to govern.
Risk Score Triggers Wrong Escalation
A risk score, severity label, confidence value, or threshold result triggers the wrong escalation path.
Risk Signal Escalates Beyond Evidence
A risk signal, warning, score, or concern escalates farther than the available evidence supports.
Routing Overrides Task Intent
Routing, mode selection, agent behavior, or workflow classification sends the task down a path that overrides what the user was trying to accomplish.
Saved Reference No Longer Works
A saved source, citation, file reference, prompt reference, or workflow pointer previously worked but no longer resolves to the expected object or meaning.
Severity Increases Without New Evidence
The severity, risk, confidence, or escalation level increases even though no new evidence has been added.
Small Change Produces Large Downstream Effects
A small prompt, schema, policy, output, or workflow change creates unexpectedly large effects in downstream steps.
Small Issue Keeps Escalating
A small issue, warning, uncertainty, or correction keeps increasing in severity, scope, or workflow impact across later steps.
Too Many Tool Calls
The AI or agent makes more tool calls, searches, retrievals, API calls, or integration actions than the task requires or permits.
Tool Call Contract Mismatch
The AI or agent calls a tool with names, arguments, types, modes, or shapes that do not match the declared tool interface.
Version Change Breaks Existing Prompt
A prompt that previously produced usable results stops working after a version change in the model, tool, policy, schema, product surface, or workflow.
Wrong Field Types
The AI returns fields with values whose types do not match the expected schema, such as strings where numbers, booleans, arrays, objects, or enums are required.