PENGO Workbench
Workbench / Lenses / Invariant Lens

LEN-0210

Invariant Lens

Verifies that declared invariants are structurally enforceable and not contradicted by observed structure.

Primary Pattern Matches

PAT-0330

Invariant Breakage

A structural condition where an observed state violates a declared invariant that is supposed to remain true.

Lens Application

The Invariant Lens is a primary inspection mechanism for Invariant Breakage because it compares declared invariants with the structural conditions that are supposed to preserve them. It helps surface places where enforcement is missing, inconsistent, or contradicted by observed states without treating the finding as a final verdict.

Inspect For

  • Declared invariants contradicted by observed state
  • Missing structural enforcement of invariant conditions
  • Exceptions or transitions that bypass invariant protection
  • Conflicting rules that make the invariant unstable

Avoid

Treat a contradiction as an inspection signal rather than conclusive evidence without checking whether scope, timing, or enforcement boundaries affect the observation.

PAT-0160

Schema Breakage

A structural condition where an instance, graph, payload, or object violates the type, shape, or rule requirements of a declared schema.

Lens Application

The Invariant Lens is a primary inspection mechanism for Schema Breakage because it compares declared schema invariants with the actual structure being observed. It helps surface places where required fields, types, cardinalities, relationships, or validation rules are structurally contradicted or cannot be consistently enforced.

Inspect For

  • Required fields or nodes that are missing
  • Type, shape, or cardinality mismatches
  • Constraints contradicted by observed structure
  • Invariants that validation cannot reliably enforce

Avoid

Treat structural mismatches as inspection signals rather than confirmation that Schema Breakage has been established.

PAT-0390

Threshold Breach

A structural condition where an observed metric, value, state, or condition exceeds a declared quantitative or qualitative threshold.

Lens Application

The Invariant Lens is a primary inspection mechanism for Threshold Breach because it examines whether the threshold functions as a true invariant or only an informal expectation. It helps surface mismatches between declared constraints and observed states that exceed, bypass, weaken, or contradict those constraints.

Inspect For

  • Declared limits, ranges, or forbidden states
  • Observed values or conditions beyond those limits
  • Enforcement gaps or bypass paths
  • Conflicts between invariant claims and system behavior

Avoid

Treat an exceeded value as an inspection signal rather than confirmation that Threshold Breach has occurred without checking whether the threshold was structurally declared, enforceable, and relevant to the observed condition.

Secondary Pattern Matches

PAT-0100

Authority Collision

A structural condition where multiple authorities claim governance over the same region without a declared precedence or resolution rule.

Lens Application

The Invariant Lens applies indirectly to Authority Collision by examining whether required governance invariants still hold when multiple authorities govern the same region. It can surface stress points where enforcement assumptions, precedence expectations, or consistency requirements are undermined by unresolved authority overlap.

Inspect For

  • Invariants depending on a single governing authority
  • Conflicting rules applied to the same governed region
  • Missing precedence or escalation conditions
  • Enforcement paths blocked by overlapping authority claims

Avoid

Treat invariant tension as a supporting inspection signal, not confirmation that Authority Collision exists.

Related Issues

ISS-0064

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.

ISS-0115

Agent Gets Conflicting Tool Authority

An agent receives conflicting authority signals about whether, when, or how it may use a tool, connector, function, or integration.

ISS-0026

AI Output Breaks Parser

The AI output causes a parser, validator, importer, or structured-output consumer to fail.

ISS-0088

Approval Depends on Output That Needs Approval

A required approval depends on an AI output or workflow result that itself cannot be produced or trusted until approval is granted.

ISS-0035

Conflicting Instructions From Different Authorities

Instructions from different sources, roles, policies, prompts, tools, or workflow authorities conflict without a clear rule for which one governs.

ISS-0097

Downstream Steps Magnify Hallucinated Claim

A hallucinated or unsupported claim from an AI output is reused by later workflow steps until it becomes more influential than the evidence supports.

ISS-0027

Duplicate Fields With Same Meaning

The AI returns multiple fields, labels, sections, or structured elements that carry the same meaning and create ambiguity about which one should be used.

ISS-0066

Duplicate Output Sections

The AI repeats sections, headings, blocks, or output areas in a way that creates redundancy, confusion, or downstream handling problems.

ISS-0028

Format Rule Too Weak

The format instruction is too vague, incomplete, or optional to reliably produce output that satisfies the expected structure.

ISS-0023

Hallucinated Fields

The AI adds fields, keys, attributes, columns, or structured elements that were not declared, requested, or allowed by the expected schema.

ISS-0039

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.

ISS-0036

Human Review and Automation Disagree

A human review result and an automated AI or workflow result disagree without a declared rule for resolving the difference.

ISS-0021

Invalid JSON Output

The AI returns malformed JSON or structured output that cannot be parsed.

ISS-0022

Missing Required Fields

The AI returns structured output that omits fields required by the schema, workflow, parser, form, or downstream consumer.

ISS-0025

Nested Fields Do Not Match

The AI returns nested structured fields whose internal shape, hierarchy, parent-child relationship, or contained values do not match the expected structure.

ISS-0018

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.

ISS-0060

Output Exceeds Length Limit

The AI output exceeds a declared length, token, word, character, section, field, or size limit.

ISS-0083

Parallel Reviews Never Agree

Parallel AI, human, workflow, or tool reviews keep producing different results without resolving into a shared decision state.

ISS-0016

Permissions Conflict After Being Combined

Permissions, approvals, roles, policies, or authority rules that seem valid separately conflict when combined in the same workflow or AI action.

ISS-0090

Policy Decision Depends on Itself

A policy decision requires the outcome of the same policy decision before it can be made.

ISS-0065

Risk Score Triggers Wrong Escalation

A risk score, severity label, confidence value, or threshold result triggers the wrong escalation path.

ISS-0095

Risk Signal Escalates Beyond Evidence

A risk signal, warning, score, or concern escalates farther than the available evidence supports.

ISS-0037

Same Case Has Conflicting Policies

The same case appears to be governed by multiple policies, rules, or standards that point to incompatible outcomes.

ISS-0091

Schema Reference Loops Without Base Case

A schema, field, type, object, or structured reference points through a loop without a base case that allows validation or interpretation to resolve.

ISS-0102

Severity Increases Without New Evidence

The severity, risk, confidence, or escalation level increases even though no new evidence has been added.

ISS-0107

Single Field Carries Too Many Obligations

One field, label, score, status, or structured value is expected to carry too many meanings, decisions, or workflow obligations.

ISS-0098

Small Issue Keeps Escalating

A small issue, warning, uncertainty, or correction keeps increasing in severity, scope, or workflow impact across later steps.

ISS-0017

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.

ISS-0110

Tool Rules and Prompt Rules Conflict

Tool, connector, function, or MCP rules conflict with prompt instructions, causing the AI or agent to face incompatible requirements.

ISS-0006

Validation Result Changes on Retry

A validation, grading, review, classification, or pass/fail result changes after retry even though the input and declared validation rules did not change.

ISS-0024

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.