ci, ctb: Reduce devnet sequencer drift

op-node: Monitor static peers in a background process

fix: wrong arg name

op-chain-ops: Ignore messages from sources other than the L2XDM

Co-authored-by: protolambda <proto@protolambda.com>

feat(fd): support Bedrock networks

op-node: Properly check storage proof

The storage proof was not being properly checked. This PR corrects the check, and adds unit/fuzz tests around the storage proof checks to prevent regressions.

Fixes CLI-3344

Currently, connections with static peers are not re-established in the event of disconnection. This causes us to have to restart replicas whenever we restart the sequencer. This PR adds a background process to monitor static peers for connectedness, and automatically reconnect when necessary.

The `relayMessage` function on the message passer is public. To guard against spoofed withdrawals, we need to check the message's source to ensure that only messages from the L2XDM are included in the withdrawals migration.

Fixes CLI-3337
Fixes CLI-3331

docs: Add note about small contributions

I saw this in the foundry docs and thought it would be good to add to ours

The sequencer drift was larger than the sequencing window. This can cause reorgs. To fix I reduced the drift to 10s, which is lower than the 45 second sequencing window specified via the L1 block time and the window size.

contracts-bedrock: add TransferOnion contract

Adds the `TransferOnion` contract which can be used to batch
send tokens. The set of transfers are committed to during
contract initialization, and the token owner must approve
the `TransferOnion` contract and then anybody can unwrap
the commitment to the transfers.

The transfers are committed to with a hash onion. This is
a simple vector commitment that must be linearly unwrapped.
Each hash commits to the next hash and the very first hash
is `bytes32(0)`.

fix(specs): Use alias syntax to address long lines

Add internal certman package with mTLS certificate reload fixes for k8s volumes

Updates the fault detector to support Bedrock networks. Bedrock support
is slightly hacky because we still also need to support the legacy
system until mainnet has been upgraded to Bedrock. We will want to
update the fault detector again once mainnet is upgraded to simplify the
service and remove the legacy components.

op-batcher: Watch timeouts [ENG-2625]

Update bridges.md