test(ci): skip e2e_multi_validator_node_key_store entirely on merge-train/spartan

[AztecBot]

Summary

PR #23344 was dequeued again at https://github.com/AztecProtocol/aztec-packages/actions/runs/26370196367. The flake-flag added in #23528 (error_regex: "Tx dropped by P2P node") wasn't enough — ci3 retries once and the multi-validator web3signer suite has now been seen failing in a second pipelining-related mode that the regex doesn't match.

Both observed modes hit the same should build blocks & attest with multiple validator keys case:

• Tx dropped by P2P node at Promise.all index 0 (the original mode, ci.aztec-labs.com/9a5fa90aa18f62e7).
• Proposer misses the slot's 5-attestation deadline → AttestationTimeoutError: Timeout collecting attestations for slot 29: 0/5, followed by Simulation error: Block 0x… not found in the node. This might indicate a reorg has occurred. (ci.aztec-labs.com/b91d3218b5e88ae4).

Following the same pattern as #23542 for e2e_ha_full: collapse the existing error_regex entry into a single skip: true so the suite is outright skipped on merge-train/spartan until proposer pipelining stabilises.

Notes

• Targets merge-train/spartan per yarn-project default base.
• Owner kept as @PaLLa (current owner of the entry from test: flag e2e_multi_validator_node_key_store P2P tx-dropped failure as flake #23528).

Requested in Slack by @PhilWindle.

---

Created by claudebox · group: slackbot