docs: threat model for node#24465
Merged
Merged
Conversation
Adds a new `THREAT_MODEL.md` doc. This document describes the threat model of the Aztec L2 network: how transactions flow from users into the proven chain, what each participant can and cannot do, and the properties the implementation must uphold. It is intended as a guideline for the security of the node implementation (`yarn-project`) and its interaction with the L1 rollup contracts (`l1-contracts`). **In scope**: transaction dissemination and the mempool, the p2p layer, block and checkpoint production, committee attestation, L1 checkpoint submission and sync, epoch proving, slashing, and the escape hatch. **Out of scope**: client-side private execution and proving (PXE, wallets), hardening of a node's public RPC interface, the cryptographic soundness of the proving system itself (treated as an assumption below), and L1 governance internals.
fcarreiro
reviewed
Jul 2, 2026
| | Malicious proposer | Censor txs; waste its own slot and the next one (pipelining); post an unattested/badly-attested checkpoint to L1; equivocate | Convince honest nodes of an invalid state transition; corrupt state beyond slot N+1; brick other nodes' sync | Re-execution before attestation/adoption; pipeline depth capped at 2; delayed attestation verification + permissionless invalidation; slashing (§5) | | ||
| | Committee minority (< quorum, not proposer) | Withhold their own attestations; equivocate (slashable) | Prevent a slot from being attested; get honest members slashed | Quorum only needs ⌊2n/3⌋+1 of n; equivocation slashing | | ||
| | Committee quorum (≥ ⌊2n/3⌋+1 malicious) | Post invalid-but-attested checkpoints that nodes follow blindly; withhold tx data; halt the pending chain until the epoch prune | Get invalid state proven or exited on L1 (A2); avoid slashing for data withholding / attested-invalid | Unprovable state → prune at proof-window expiry; archiver preemptive unwind; slashing incl. `DATA_WITHHOLDING` | | ||
| | Malicious prover | Nothing by proving (proofs are verified); grief by *not* proving | Prove an invalid transition (A2); steal fees/rewards via forged proof calldata | Proof verification; fees bound to attested headers (see §4.4); A4 for liveness | |
Contributor
There was a problem hiding this comment.
I know we've talked about this a thousand times (maybe look for "training wheels" on Slack). Isn't there a mechanism today (but IIRC we don't want it in the future) where the committee/some part of the system would still not accept a new state if it differs from what the validators saw?
Contributor
Author
There was a problem hiding this comment.
The committee can be seen as "training wheels" for the proving system, since a state transition cannot be deemed valid in L1 unless it's both proven and attested to. I've added a mention of that now.
fcarreiro
approved these changes
Jul 7, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Adds a new
THREAT_MODEL.mddoc. This document describes the threat model of the Aztec L2 network: how transactions flow from users into the proven chain, what each participant can and cannot do, and the properties the implementation must uphold. It is intended as a guideline for the security of the node implementation (yarn-project) and its interaction with the L1 rollup contracts (l1-contracts).In scope: transaction dissemination and the mempool, the p2p layer, block and checkpoint production, committee attestation, L1 checkpoint submission and sync, epoch proving, slashing, and the escape hatch.
Out of scope: client-side private execution and proving (PXE, wallets), hardening of a node's public RPC interface, the cryptographic soundness of the proving system itself (treated as an assumption below), and L1 governance internals.