Where does technical capability end — and real execution authority begin?
Choose one real effect: a payment, data release, agent action, administrative mutation or publication. The following ten questions show whether a robust, current and evidence-bound execution boundary exists between identity and effect.
IDENTITY
AUTHORITY
STATE
EXECUTION
EVIDENCE
WHAT THIS IS ABOUT
Not an audit badge. A robust gap map.
This assessment does not replace technical due diligence, certification, legal advice or independent acceptance. It translates the architecture question into plain language: Can this exact effect execute now — and why? For each point, mark whether your architecture clearly closes the boundary, needs review, or remains unknown.
Claim Boundary: A self-assessment does not prove security or legal compliance. It only identifies execution boundaries that merit further technical or institutional review.
GAP MAP
You are not checking features. You are checking control of effect.
IDENTITY
TIME
AGENT
STATE
SURFACE
OVERRIDE
EVIDENCE
1 · Identity ≠ Authority
Does a successful login suffice to trigger a critical effect?
A login answers who or what authenticated. It does not automatically answer whether this specific action is legitimate and institutionally authorized right now.
WHAT TO LOOK AT
Check whether an independent authority decision exists between authentication and productive effect.
WHAT A ROBUST BOUNDARY DOES
A robust boundary binds identity, concrete action, target, purpose and current authority separately.
IDENTITY → AUTHORITY → EFFECT
2 · Continuing Authority
Is authority revalidated immediately before the point of effect?
An approval may have been valid and become invalid seconds later. Mandates can be revoked, state can change and dependencies can expire.
WHAT TO LOOK AT
Find the last control point before settlement, mutation, release or publication. Is an old PASS simply reused?
WHAT A ROBUST BOUNDARY DOES
A robust boundary revalidates effect-relevant state immediately before execution.
PASS(t-3m) ≠ RIGHT(t0)
3 · Agentic Execution
Can an AI agent directly trigger a productive tool call?
Models can propose, plan and request tools. None of that creates institutional authority. Prompt injection, bad context or hallucination must not become productive effect by themselves.
WHAT TO LOOK AT
Trace an agent tool call to the real effect: can model output alone force the mutation?
WHAT A ROBUST BOUNDARY DOES
Execution must remain outside the model and bound to deterministic authority, state and scope conditions.
MODEL REQUEST ⇏ EXECUTION RIGHT
4 · State Reality
Can external data or sensors directly create authority?
APIs, telemetry and sensors provide observations. Observation may invalidate execution, but it is not automatically an authority source itself.
WHAT TO LOOK AT
Check provenance, freshness, source independence and handling of conflicting sources.
WHAT A ROBUST BOUNDARY DOES
A robust boundary does not silently turn unverified or stale state into PASS.
OBSERVATION ≠ AUTHORITY
5 · Atomic Consumption
Can the same approval or capability be used more than once?
A valid token becomes dangerous if it can be replayed or consumed in parallel. Async APIs make race conditions easy to miss.
WHAT TO LOOK AT
Check nonces, JTI/consumption state, atomic writes and parallel requests.
WHAT A ROBUST BOUNDARY DOES
A consumable right must be uniquely bound and atomically consumable.
ONE RIGHT → ONE EFFECT
6 · Execution Surface
Are all productive effect paths actually covered?
The strongest control layer is irrelevant if a forgotten endpoint, queue, admin tool or legacy route can trigger the same effect outside the boundary.
WHAT TO LOOK AT
Inventory APIs, workers, queues, webhooks, admin and recovery paths that can mutate real state.
WHAT A ROBUST BOUNDARY DOES
Control is robust only across execution surfaces that are actually inventoried and controlled.
UNREGISTERED PATH → REVIEW
7 · Override Accountability
What happens when a BLOCK must be overridden in an emergency?
An override must not delete or reinterpret the original BLOCK. Otherwise the most important forensic fact disappears.
WHAT TO LOOK AT
Check break-glass, supervisor and emergency procedures: does the original state remain intact?
WHAT A ROBUST BOUNDARY DOES
A robust override is a new, separately grounded process with its own authority and evidence.
BLOCK + NEW AUTHORITY → NEW RECEIPT
8 · Evidence
Can you later reconstruct why this exact effect was allowed?
Traditional logs often show events. Strong reconstruction requires input, authority state, rule version, system state, decision and execution to remain connected.
WHAT TO LOOK AT
Check whether evidence can be changed later and whether the rule and authority state can be reconstructed.
WHAT A ROBUST BOUNDARY DOES
Receipts can strengthen integrity and causal traceability. They do not replace judicial or regulatory evidentiary assessment.
INPUT → RULE → VERDICT → RECEIPT
9 · Cross-Domain Closure
Can a PASS in one domain neutralize a BLOCK in another?
Regulated actions often depend on several mandatory conditions at once. A positive partial result must not compensate for a mandatory exclusion.
WHAT TO LOOK AT
Check how compliance, authority and risk domains are composed.
WHAT A ROBUST BOUNDARY DOES
For mandatory conditions, closure must remain conjunctive: unresolved or BLOCK remains effective.
PASS + BLOCK = BLOCK
10 · Dependencies
What happens when a critical dependency fails or becomes stale?
A sanctions feed, authority directory or external register can fail. The key question is whether the system still executes or treats the state as unresolved.
WHAT TO LOOK AT
Check TTL, freshness, degraded mode and fail-open behavior.
WHAT A ROBUST BOUNDARY DOES
Where freshness is mandatory, stale or unavailable state must not silently produce PASS.
STALE STATE ≠ PASS
Your gap map No points marked yet.
NEXT STEP
Choose one effect. We trace it to the point of effect.
A questionnaire is not enough for technical assurance. The next step is a scoped review of one concrete effect chain: request → authority → state → execution → evidence.