feat(work-unit): volunteer-mesh verification — trust untrusted results by quorum - #25
Merged
Merged
Conversation
…s by quorum
The piece the compute plane was missing, and the core of the volunteer-mesh design. A Grant
proves WHO ran a Work Unit; it cannot prove the RESULT is correct. On a Folding@home-scale
grid (400k anonymous, churny, possibly-malicious workers) you must VERIFY results — that's
how BOINC/Folding@home actually work.
tools/work_unit.py:
- mint_work_unit: a content-addressed Work Unit (wu_id from the task def; dedup + repro),
carrying proof_mode + sandbox + ttl + replication factor.
- Verifier.redundant_quorum: run the same WU on N independent workers; accept only the output
a quorum agrees on (identical digest). No quorum -> REJECTED, never accepted unverified.
Flags the disagreers. Fail-closed.
- Verifier.spot_check: canary with a known answer for stochastic tasks; wrong canary -> reject.
- Reputation: per-worker verified-success record; weights allocation, routes around bad actors
(a liar earns nothing). trusted() gates on score + min samples.
- verify_and_score: quorum-verify + update reputation in one call.
Implements the Dual-Orchestration Model PROOF_MODE (redundant|spot_check|tee|zk|optimistic) and
the CluBORun "reliable cluster as reference verifier" idea. Distilled from the volunteer-computing
corpus (BOINC OOBE, agent-first node, Urbit/CoreOS, CluBORun, dual-orchestration).
capd/volunteer-mesh-verification.mesh.capd.json (caps.compute.volunteer-mesh-verification) —
ingested by the commons as reproducible. Tests: +8 = 118 tools tests green. Makefile `verify`.
Corpus synthesis (docs/VOLUNTEER_MESH.md, the full superset design) lands next after the PDF review.
This was referenced Aug 4, 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
1 participant
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.
The piece the compute plane was missing, and the core of the volunteer-mesh design. A Grant proves who ran a Work Unit; it cannot prove the result is correct. On a Folding@home-scale grid (400k anonymous, churny, possibly-malicious workers) you must verify results — how BOINC/Folding@home actually work.
tools/work_unit.pywu_idfrom the task def → dedup + repro), carryingproof_mode+sandbox+ttl+ replication.Implements the Dual-Orchestration
PROOF_MODE(redundant|spot_check|tee|zk|optimistic) + the CluBORun reference-verifier idea. Distilled from your volunteer-computing corpus.118 tools tests green.
capd/volunteer-mesh-verificationingested by the commons as reproducible.The full corpus synthesis (
docs/VOLUNTEER_MESH.md— the superset design) lands next after the PDF review completes.