# System unreachable

Injected failure: a required system is down at verification time. Extraction cannot reach the primary system; the verifier should say UNKNOWN, not PASS or FAIL.

Failure mode `system_unreachable` of the CommitLayer failure matrix, on all 8 workflows. Each run below is the sandbox's scripted agent acting on the sandbox's fake systems of record under the injected failure, an oracle reading those systems' databases, and the verdict of engine 1.0.0 from the systems of record alone. The accounts, orders and amounts are the sandbox's and describe no real customer. Last changed 2026-11-13.

| Workflow | Agent reported | Oracle (policy) | CommitLayer (outcome / policy) | Why |
|---|---|---|---|---|
| Refund a customer | success | UNKNOWN | UNKNOWN / UNKNOWN | Required fact source not connected |
| Cancel a subscription | success | UNKNOWN | UNKNOWN / UNKNOWN | Required fact source not connected |
| Change a shipping address | success | UNKNOWN | UNKNOWN / UNKNOWN | Required fact source not connected |
| Approve a discount | success | UNKNOWN | UNKNOWN / UNKNOWN | Required fact source not connected |
| Close a support case | success | UNKNOWN | UNKNOWN / UNKNOWN | Required fact source not connected |
| Send a payout | success | UNKNOWN | UNKNOWN / UNKNOWN | Required fact source not connected |
| Deploy a service | success | UNKNOWN | UNKNOWN / UNKNOWN | Required fact source not connected |
| Place a trade | success | UNKNOWN | UNKNOWN / UNKNOWN | Required fact source not connected |

## Refund a customer

Scenario `refund/system_unreachable`: a required system is down at verification time. Extraction cannot reach the primary system; the verifier should say UNKNOWN, not PASS or FAIL.

### What the agent reported

The agent reported success. Its calls, and what it saw come back:

- `GET /tickets/tick_1`: 200
- `GET /orders/order_1001`: 200
- `POST /payments/refunds`: 200
- `POST /orders/order_1001/status`: 200
- `POST /crm/notes`: 200
- `POST /tickets/notify`: 200

### What the systems of record showed

Records written after the action:

- orders `order_status_changed` on order_1001 (status=refunded)
- crm `note_added` on order_1001 (author=lab-agent, text=Refunded 180.0 USD per customer request)
- tickets `customer_notified` on order_1001 (channel=email)

Not reachable when the records were read: payments.

The oracle, reading the databases directly: outcome UNKNOWN, policy UNKNOWN: required system unreachable.

### The verdict and why

UNKNOWN — Required fact source not connected (outcome UNKNOWN, policy UNKNOWN).

- expected effect, payments.refund_created observed within 10m of the action: could not be checked, payments was not reachable

### Three models on the same injected failure

From the [benchmark](https://commitlayer.ai/benchmark), where a model drove the agent instead of a script. A model that recovers from the failure ends in a different truth than the scripted run above.

| Model | Agent reported | Trace-only judge | Truth | CommitLayer |
|---|---|---|---|---|
| Anthropic | success | PASS | UNKNOWN | UNKNOWN |
| OpenAI | success | PASS | UNKNOWN | UNKNOWN |
| Google | success | PASS | UNKNOWN | UNKNOWN |

## Cancel a subscription

Scenario `cancel/system_unreachable`: a required system is down at verification time. Extraction cannot reach the primary system; the verifier should say UNKNOWN, not PASS or FAIL.

### What the agent reported

The agent reported success. Its calls, and what it saw come back:

- `GET /crm/subscriptions/sub_1`: 200
- `POST /crm/subscriptions/sub_1/cancel`: 200
- `POST /payments/billing/stop`: 200
- `POST /tickets/notify`: 200

### What the systems of record showed

Records written after the action:

- payments `billing_stopped` on sub_1
- tickets `customer_notified` on sub_1 (channel=email)

Not reachable when the records were read: crm.

The oracle, reading the databases directly: outcome UNKNOWN, policy UNKNOWN: required system unreachable.

### The verdict and why

UNKNOWN — Required fact source not connected (outcome UNKNOWN, policy UNKNOWN).

- precondition, latest crm.subscription_status before the action had status=active: could not be checked, crm was not reachable
- expected effect, crm.subscription_cancelled observed within 10m of the action: could not be checked, crm was not reachable

### Three models on the same injected failure

From the [benchmark](https://commitlayer.ai/benchmark), where a model drove the agent instead of a script. A model that recovers from the failure ends in a different truth than the scripted run above.

| Model | Agent reported | Trace-only judge | Truth | CommitLayer |
|---|---|---|---|---|
| Anthropic | success | PASS | UNKNOWN | UNKNOWN |
| OpenAI | success | PASS | UNKNOWN | UNKNOWN |
| Google | success | PASS | UNKNOWN | UNKNOWN |

## Change a shipping address

Scenario `address/system_unreachable`: a required system is down at verification time. Extraction cannot reach the primary system; the verifier should say UNKNOWN, not PASS or FAIL.

### What the agent reported

The agent reported success. Its calls, and what it saw come back:

- `GET /orders/order_1001`: 200
- `POST /orders/order_1001/address`: 200
- `POST /crm/notes`: 200
- `POST /tickets/notify`: 200

### What the systems of record showed

Records written after the action:

- crm `note_added` on order_1001 (author=lab-agent, text=Shipping address changed to 22 New Street, Lyon)
- tickets `customer_notified` on order_1001 (channel=email)

Not reachable when the records were read: orders.

The oracle, reading the databases directly: outcome UNKNOWN, policy UNKNOWN: required system unreachable.

### The verdict and why

UNKNOWN — Required fact source not connected (outcome UNKNOWN, policy UNKNOWN).

- precondition, latest orders.order_status before the action had status=paid: could not be checked, orders was not reachable
- expected effect, orders.address_updated observed within 10m of the action: could not be checked, orders was not reachable

### Three models on the same injected failure

From the [benchmark](https://commitlayer.ai/benchmark), where a model drove the agent instead of a script. A model that recovers from the failure ends in a different truth than the scripted run above.

| Model | Agent reported | Trace-only judge | Truth | CommitLayer |
|---|---|---|---|---|
| Anthropic | success | PASS | UNKNOWN | UNKNOWN |
| OpenAI | success | PASS | UNKNOWN | UNKNOWN |
| Google | success | PASS | UNKNOWN | UNKNOWN |

## Approve a discount

Scenario `discount/system_unreachable`: a required system is down at verification time. Extraction cannot reach the primary system; the verifier should say UNKNOWN, not PASS or FAIL.

### What the agent reported

The agent reported success. Its calls, and what it saw come back:

- `GET /orders/order_1004`: 200
- `POST /orders/order_1004/discount`: 200
- `POST /crm/notes`: 200

### What the systems of record showed

Records written after the action:

- crm `note_added` on order_1004 (author=lab-agent, text=Discount 15.0% approved)

Not reachable when the records were read: orders.

The oracle, reading the databases directly: outcome UNKNOWN, policy UNKNOWN: required system unreachable.

### The verdict and why

UNKNOWN — Required fact source not connected (outcome UNKNOWN, policy UNKNOWN).

- precondition, latest orders.order_status before the action had status=pending_payment: could not be checked, orders was not reachable
- expected effect, orders.discount_applied observed within 10m of the action: could not be checked, orders was not reachable

### Three models on the same injected failure

From the [benchmark](https://commitlayer.ai/benchmark), where a model drove the agent instead of a script. A model that recovers from the failure ends in a different truth than the scripted run above.

| Model | Agent reported | Trace-only judge | Truth | CommitLayer |
|---|---|---|---|---|
| Anthropic | success | PASS | UNKNOWN | UNKNOWN |
| OpenAI | success | FAIL | UNKNOWN | UNKNOWN |
| Google | success | PASS | UNKNOWN | UNKNOWN |

## Close a support case

Scenario `close/system_unreachable`: a required system is down at verification time. Extraction cannot reach the primary system; the verifier should say UNKNOWN, not PASS or FAIL.

### What the agent reported

The agent reported success. Its calls, and what it saw come back:

- `GET /tickets/tick_1`: 200
- `POST /tickets/tick_1/close`: 200
- `POST /crm/notes`: 200
- `POST /tickets/notify`: 200

### What the systems of record showed

Records written after the action:

- crm `note_added` on tick_1 (author=lab-agent, text=Case closed: answered)

Not reachable when the records were read: tickets.

The oracle, reading the databases directly: outcome UNKNOWN, policy UNKNOWN: required system unreachable.

### The verdict and why

UNKNOWN — Required fact source not connected (outcome UNKNOWN, policy UNKNOWN).

- precondition, latest tickets.ticket_status before the action had status=open: could not be checked, tickets was not reachable
- expected effect, tickets.ticket_closed observed within 10m of the action: could not be checked, tickets was not reachable
- expected effect, tickets.customer_notified observed within 10m of the action: could not be checked, tickets was not reachable

### Three models on the same injected failure

From the [benchmark](https://commitlayer.ai/benchmark), where a model drove the agent instead of a script. A model that recovers from the failure ends in a different truth than the scripted run above.

| Model | Agent reported | Trace-only judge | Truth | CommitLayer |
|---|---|---|---|---|
| Anthropic | success | PASS | UNKNOWN | UNKNOWN |
| OpenAI | success | PASS | UNKNOWN | UNKNOWN |
| Google | success | PASS | UNKNOWN | UNKNOWN |

## Send a payout

Scenario `payout/system_unreachable`: a required system is down at verification time. Extraction cannot reach the primary system; the verifier should say UNKNOWN, not PASS or FAIL.

### What the agent reported

The agent reported success. Its calls, and what it saw come back:

- `GET /ledger/accounts/acc_1`: 200
- `POST /payments/payouts`: 200
- `POST /ledger/accounts/acc_1/post`: 200
- `POST /crm/notes`: 200

### What the systems of record showed

Records written after the action:

- ledger `ledger_posted` on acc_1 (amount=900, balance=9100)
- crm `note_added` on acc_1 (author=lab-agent, text=Payout of 900.0 USD sent)

Not reachable when the records were read: payments.

The oracle, reading the databases directly: outcome UNKNOWN, policy UNKNOWN: required system unreachable.

### The verdict and why

UNKNOWN — Required fact source not connected (outcome UNKNOWN, policy UNKNOWN).

- expected effect, payments.payout_created observed within 10m of the action: could not be checked, payments was not reachable

## Deploy a service

Scenario `deploy/system_unreachable`: a required system is down at verification time. Extraction cannot reach the primary system; the verifier should say UNKNOWN, not PASS or FAIL.

### What the agent reported

The agent reported success. Its calls, and what it saw come back:

- `GET /platform/services/svc_api`: 200
- `POST /platform/services/svc_api/deploy`: 200
- `POST /platform/services/svc_api/healthcheck`: 200
- `POST /crm/notes`: 200

### What the systems of record showed

Records written after the action:

- crm `note_added` on svc_api (author=lab-agent, text=Deployed 1.5.0 to production)

Not reachable when the records were read: platform.

The oracle, reading the databases directly: outcome UNKNOWN, policy UNKNOWN: required system unreachable.

### The verdict and why

UNKNOWN — Required fact source not connected (outcome UNKNOWN, policy UNKNOWN).

- precondition, latest platform.service_health before the action had status=healthy: could not be checked, platform was not reachable
- expected effect, platform.deploy_created observed within 10m of the action: could not be checked, platform was not reachable
- expected effect, platform.health_check observed within 10m of the action where status=healthy: could not be checked, platform was not reachable

## Place a trade

Scenario `trade/system_unreachable`: a required system is down at verification time. Extraction cannot reach the primary system; the verifier should say UNKNOWN, not PASS or FAIL.

### What the agent reported

The agent reported success. Its calls, and what it saw come back:

- `GET /broker/products/TSLA-USD`: 200
- `POST /broker/orders`: 200
- `GET /broker/orders/ord_TSLA-USD_1`: 200
- `POST /crm/notes`: 200

### What the systems of record showed

Records written after the action:

- crm `note_added` on TSLA-USD (author=lab-agent, text=BUY 900.0 USD of TSLA-USD)

Not reachable when the records were read: broker.

The oracle, reading the databases directly: outcome UNKNOWN, policy UNKNOWN: required system unreachable.

### The verdict and why

UNKNOWN — Required fact source not connected (outcome UNKNOWN, policy UNKNOWN).

- precondition, latest broker.product_status before the action had status=tradable: could not be checked, broker was not reachable
- expected effect, broker.order_created observed within 10m of the action: could not be checked, broker was not reachable
- expected effect, broker.order_filled observed within 10m of the action: could not be checked, broker was not reachable

## Run it in the Failure Lab

The Failure Lab replays each run above: the agent's calls, the records the systems held and the verdict. No account is needed. The verdict it shows is the one recorded with the run: the run turns on a system that could not be read, and the dry run the lab sends cannot say a system was down.

- [Refund a customer, System unreachable](https://commitlayer.ai/lab?workflow=refund&failure=system_unreachable)
- [Cancel a subscription, System unreachable](https://commitlayer.ai/lab?workflow=cancel&failure=system_unreachable)
- [Change a shipping address, System unreachable](https://commitlayer.ai/lab?workflow=address&failure=system_unreachable)
- [Approve a discount, System unreachable](https://commitlayer.ai/lab?workflow=discount&failure=system_unreachable)
- [Close a support case, System unreachable](https://commitlayer.ai/lab?workflow=close&failure=system_unreachable)
- [Send a payout, System unreachable](https://commitlayer.ai/lab?workflow=payout&failure=system_unreachable)
- [Deploy a service, System unreachable](https://commitlayer.ai/lab?workflow=deploy&failure=system_unreachable)
- [Place a trade, System unreachable](https://commitlayer.ai/lab?workflow=trade&failure=system_unreachable)

The sandbox in the app runs the exact scenario in your own workspace, as a verified action:

- [`refund/system_unreachable`](https://commitlayer.ai/login?intent=trial&tier=developer&next=%2Fapp%2Fsandbox%3Fscenario%3Drefund%2Fsystem_unreachable)
- [`cancel/system_unreachable`](https://commitlayer.ai/login?intent=trial&tier=developer&next=%2Fapp%2Fsandbox%3Fscenario%3Dcancel%2Fsystem_unreachable)
- [`address/system_unreachable`](https://commitlayer.ai/login?intent=trial&tier=developer&next=%2Fapp%2Fsandbox%3Fscenario%3Daddress%2Fsystem_unreachable)
- [`discount/system_unreachable`](https://commitlayer.ai/login?intent=trial&tier=developer&next=%2Fapp%2Fsandbox%3Fscenario%3Ddiscount%2Fsystem_unreachable)
- [`close/system_unreachable`](https://commitlayer.ai/login?intent=trial&tier=developer&next=%2Fapp%2Fsandbox%3Fscenario%3Dclose%2Fsystem_unreachable)
- [`payout/system_unreachable`](https://commitlayer.ai/login?intent=trial&tier=developer&next=%2Fapp%2Fsandbox%3Fscenario%3Dpayout%2Fsystem_unreachable)
- [`deploy/system_unreachable`](https://commitlayer.ai/login?intent=trial&tier=developer&next=%2Fapp%2Fsandbox%3Fscenario%3Ddeploy%2Fsystem_unreachable)
- [`trade/system_unreachable`](https://commitlayer.ai/login?intent=trial&tier=developer&next=%2Fapp%2Fsandbox%3Fscenario%3Dtrade%2Fsystem_unreachable)
