Trigger and oracle

STALL's trigger is the Chainlink L2 Sequencer Uptime Feed on Base. The feed reports whether the Base sequencer is up or down, and when its status last changed. STALL is designed to read nothing else: no claim, no loss report and no committee decision goes into a payout.

The Sequencer Uptime Feed#

The feed exposes the standard Chainlink aggregator interface. Status changes are recorded as rounds, and latestRoundData() returns the most recent one.

Reading the feed#

A round with answer of 1 marks the start of a halt. The next round with answer of 0 marks its end.

Measuring halt duration#

In the intended design, STALL measures a halt from the feed's own status timestamps:

1. The halt starts at the status timestamp of the round that reports the sequencer down.
2. The halt ends at the status timestamp of the round that reports it up again.
3. The halt duration is the difference between the two.

A halt is covered if it begins during the cover period. If the halt duration is longer than a tier's threshold, every active cover position in that tier is triggered, and so is every active cover position in each shorter tier.

The measured duration is what counts. It can differ from the duration reported on a status page, and from the disruption individual users experienced. See oracle failure or late flag.

How status reaches Base#

The feed on Base is updated through messages sent from Ethereum. Sequencer status is reported on Ethereum and delivered to the feed on Base as a cross-chain message. This route matters for STALL, because a message to Base can only be executed when Base is producing blocks.

Why nothing executes during a halt#

While Base is halted, nothing executes on Base: not user transactions, not updates to the feed, and not STALL. The down and up status updates are recorded when the chain resumes. The feed is designed so that these queued status updates are processed when the sequencer resumes, and STALL measures the halt from the status timestamps recorded in those rounds.

This is why STALL pays after the restart rather than during the halt. It is also why the design does not rely on anyone watching the halt as it happens: the record of the halt reaches Base when the chain restarts, and settlement reads it from there.

Why parametric#

A parametric trigger pays on an event that can be observed on-chain, not on an assessment of loss. For STALL this means:

• No claim. Cover buyers do not submit evidence or fill in a form.
• No committee. No one votes on whether a halt counts.
• One rule for everyone. Every active cover position in a triggered tier pays its full payout size.

The cost is dependence on a single data source, and basis risk: the payout follows the feed, not the cover buyer's loss. See Risks.

Configuration and governance#

Which feed STALL reads and how its rounds are interpreted are oracle configuration. Oracle configuration is governed by $STALL stakers, together with tiers (including their thresholds) and rates. These docs do not specify quorum or timelock parameters for those decisions; see governance risk.

Reference#

Chainlink documents how these feeds work in L2 Sequencer Uptime Feeds (opens in a new tab). STALL is an independent project and is not affiliated with, or endorsed by, Chainlink.