Repository navigation
quest: finish Scoped peers prove a route key - #8022
Conversation
Quest-Claim: 891ddd011b23cbfe
#7937 implemented the lite-07 SETUP route key proof but kept the quest file to record its decision, which the routes epic and session integration already carry. Delete it and its links. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
|
The Quest check failed because the PR deletes Removed the dangling references in Task list (6/6 completed)
|
kixelated
changed the title
|
Warning Review limit reachedYou've used all free OSS reviews for now. Wait for the free limit to reset to keep reviewing this public repository. Next included review available in 5 minutes. View limit details
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. |
![pullfrog[bot]](https://avatars.githubusercontent.com/in/1768019?s=60&v=4){kind=small}
There was a problem hiding this comment.
✅ No new issues found.
Reviewed changes
- Deleted
quest/m2/cluster-routing/routes/proof.md— the quest's Goal landed in #7937 (route::proof, SETUP parameter 0x8, JSroute_key.ts, relay--cluster-route-key), and filing identities under the key belongs tosession.md. - Removed its
## Requiredentry and decision-bullet link inroutes/README.md, plus the link and## Relatedentry infederation.md— both sentences keep thesession.mdcitation and still read correctly.
I verified the PR's central claim rather than assuming it: a repo-wide grep finds no remaining reference to proof.md or "Scoped peers prove a route key", the decision survives in the epic README (Plan item 6 citing #7937, plus the "Fresh routing identity" bullet), in session.md's 2026-11-05 keying decision, and in the draft's Route Key Parameter plus rs/moq-net/src/route/proof.rs. The implementation files all exist, and deleting a finished quest matches how this tree retires other completed work.
ℹ️ Browser proof check has no quest-tree home after this
The known gap noted in the description, that no test runs a real WebTransport exportKeyingMaterial proof against the relay, was introduced by #7937 and is untouched here, so this is not a blocker. But this deletion removes the last quest-tree text that mentions browsers proving a key, and the lite-07 cut gate in quest/m2/lite07-finalize.md will no longer carry it. If that check is expected before the cut, it is worth adding there so it does not live only in PR descriptions.
deepseek-v4.1-flash-1017 (free via Pullfrog for OSS) ｜ 𝕏
Completes quest/m2/cluster-routing/routes/proof.md.
Problem
moq-dev/moq#7937 implemented this quest's Goal: the lite-07 SETUP Route Key parameter (0x8) with its exporter-bound transcript, Rust and JS proving and verification,
origin::Config::route_key, the relay's--cluster-route-keyfile, and the draft section. It also says it "Completes" the quest. Even so, it keptproof.mdto record the 2026-11-05 decision. The quest has stayed ready since, and the routes epic still lists it as Required.Approach
Delete the quest with
quest rm, dropping its Required entry inquest/m2/cluster-routing/routes/README.mdand its Related entry inquest/m2/cluster-routing/federation.md. Two prose sentences in those files also linked it. Those links are gone and each sentence keeps its other citation, Route session integration. (pullfrog[bot] pushed that fix as5e1c0c3b, after my first commit left the links dangling.)The decision is already recorded elsewhere, so deleting the file loses nothing:
session.mdcarries it in its 2026-11-05 keying decision (PeerSetup::route_key,--cluster-route-key, per-session key for unproven peers).rs/moq-net/src/route/proof.rspin the transcript.Choices made without asking
rs/moq-net/src/route/proof.rs,rs/moq-net/src/lite/setup.rs,rs/moq-relay/src/cluster.rs(--cluster-route-key), andjs/net/src/lite/route_key.ts. Filing identities under the key belongs to Route session integration, as the quest's own Goal says. Nothing in this quest's scope is left.Impact
Public API: none. Wire: none. Quest tree only.
Validation
quest checkandquest check --base origin/main: pass (578 documents).just check origin/main(nix develop): every check phase passed (quest, large_files, text, toml, biome, nix, justfile, shell, markdown, gh).test ghwas cancelled after 50 minutes. Each of its commands run alone passed, except:bash sh/local.sh python3 bench/build.test.pynever finished. It runsbench/run.sh, which takes the host-wide benchmark lock/tmp/moq-bench.lock, and another worktree'sbench/browser/run.sh origin/main --scenarios stats --rounds 5held that lock. Copies of the same test from other sessions were stuck there for about two hours. This branch changes onlyquest/*.md, and the test is byte-identical toorigin/main, so the wait is pre-existing.python3 sh/rs/test-lints.test.pyhit a 300 s cap under a load average of 45. Rerun without the cap, it passed (2 tests, 469 s).Review
Skipped by Claude Code, Claude Opus 5.5, on head
5e1c0c3b6fd79fe4487538d32399c3a00dc5d9ee: too trivial to need one. The diff deletes a quest whose implementation already landed in moq-dev/moq#7937, along with its links inquest/. It changes no code, wire, or API, andquest check --base origin/mainpasses.Follow-ups
bench/build.test.py, part of everyjust checkthroughjust gh test, takes the real host benchmark lock. It therefore blocks for as long as any benchmark session runs on the host. The test should take a lock of its own, such as one in its temporary repo, so the merge gate never queues behind a benchmark.(Written by Claude Opus 5.5)
🤖 Generated with Claude Code