3f: enrich live offer state as a first step toward cancel/reprice support

[Trynax]

Hi team,

Trynax here, I’ve had a few PRs merged in relay. I figured the team might just be busy on other things at the moment.

I only recently discovered vault-solver, and while digging into the 3F solver I noticed that the current implementation tracks live offer coverage locally, but doesn’t yet seem to track enough remote offer identity/state to support cancel/reprice lifecycle management cleanly.

From reading the current code and vendored 3F API surface, it looks like:

• the solver currently rebuilds local live-offer state from GET /v1/offer, but only stores (adapter, auction) -> expiry + principal
• createOffer currently discards the returned offer ID
• the local tracker does not retain remote offerId, nonce, status, or expectedReturn
• the solver always uses a fresh monotonic nonce for offer submission, so it is effectively opting out of the API’s “update existing mutable offer for the same (auctionId, maker, nonce)” path
• the vendored 3F API already appears to expose:
  • list offers
  • get offer by ID
  • cancel offer

Also, docs/3F-PLAN.md already calls out some related deferred lifecycle gaps:

• live offer repricing
• offer cancellation
• dynamic adapter discovery

My read is that a good first step here may be to enrich local 3F offer state before attempting actual repricing/cancel policy. Something like:

• capture/store remote offer IDs from createOffer
• track enough local state per live offer to reason about lifecycle transitions
• wire the missing cancel/get-by-id API surfaces in the local client wrapper
• expand status handling so canceled/non-covering offers are not treated as live coverage during cache rebuild
• leave actual repricing / cancel-recreate policy for a follow-up step

Before I start working on that, I wanted to sanity-check whether this matches the intended direction for the 3F solver.

Also, would love to ask: is the expected path here update-in-place via reused nonce, cancel/recreate via offer ID, or some combination of both?

[Trynax]

Bumping this with pr #71 and #93

@alrxy @oxsteins