Findings / Public reports

What we reported.
What GitHub records.

Upstream reports authored by zkasuran in the network and yellowpaper trackers. This includes work before flop-lab. Each list keeps its observation time. Follow an issue for its latest state.

flop-labs/network

Observed 2026-11-06T07:44:15.086Z. GitHub source query ↗

5 reports. 5 open at capture.

OPEN at capture / #101
gateway: flop_knowledge_manifest returns fields its output schema forbids ↗
Filed 2026-11-06T01:28:36Z. Updated 2026-11-06T01:28:37Z.
OPEN at capture / #100
docs: @flop-labs/sdk README passes response.headers to parsePaymentHeaders, which takes a Response ↗
Filed 2026-11-06T01:28:32Z. Updated 2026-11-06T01:28:33Z.
OPEN at capture / #87
gateway: flop_knowledge_get violates its advertised output schema ↗
Filed 2026-11-05T10:29:21Z. Updated 2026-11-05T17:46:42Z.
OPEN at capture / #81
feature: access form has no option for validator onboarding or devnet miners, so 8 of 17 requests do not fit it ↗
Filed 2026-11-05T06:24:34Z. Updated 2026-11-05T22:48:25Z.
OPEN at capture / #80
docs: SDK and CLI packages send bug reports to flop-core (404), and this tracker has no SDK or CLI route ↗
Filed 2026-11-05T06:24:33Z. Updated 2026-11-05T15:06:39Z.

flop-labs/yellowpaper

Observed 2026-11-06T07:44:15.721Z. GitHub source query ↗

9 reports. 0 open at capture.

CLOSED at capture / #24
erratum: §3.6 names the attestation authority the BFT committee, where the normative rule uses active validators ↗
Filed 2026-09-10T13:26:36Z. Updated 2026-09-24T17:25:53Z. Closed 2026-09-24T17:25:53Z.
CLOSED at capture / #23
erratum: Appendix A § column wrong on six rows (five emission → §2.1/§2.2, audit_fee_split_ppm → §7) ↗
Filed 2026-09-10T13:21:48Z. Updated 2026-09-24T17:25:50Z. Closed 2026-09-24T17:25:50Z.
CLOSED at capture / #22
erratum: Appendix A validators_to_eject ranks by performance, but R15.5 says ranking MUST be by stake ↗
Filed 2026-09-10T12:44:16Z. Updated 2026-10-12T15:49:22Z. Closed 2026-09-24T17:25:47Z.
CLOSED at capture / #21
erratum: §5.2/§12.1 state D-0502's withdrawal as done, but E.46 and the decision record mark it pending ↗
Filed 2026-09-10T12:44:04Z. Updated 2026-09-24T17:25:43Z. Closed 2026-09-24T17:25:43Z.
CLOSED at capture / #20
erratum: §12.4 says Shor breaks sr25519/ed25519 but R6.5d includes secp256k1 (a chain-accepted scheme) ↗
Filed 2026-09-10T12:43:53Z. Updated 2026-09-24T17:25:40Z. Closed 2026-09-24T17:25:40Z.
CLOSED at capture / #19
erratum: R9.6 tx-fee split reads 80 ppt + 10 ppt but the target is 80% + 10% (10x low) ↗
Filed 2026-09-10T12:42:43Z. Updated 2026-09-24T17:25:36Z. Closed 2026-09-24T17:25:36Z.
CLOSED at capture / #10
erratum: base-unit definition says VFY is the 10^18 base unit and also 1 VFY = 1 FLOP; the two contradict ↗
Filed 2026-09-10T08:13:05Z. Updated 2026-09-10T10:31:29Z. Closed 2026-09-10T10:31:16Z.
CLOSED at capture / #9
ambiguity: da_lease_deposit_per_byte is named per-byte but valued 200 FLOP/GB; bytes × value reads two ways (1e9x apart) ↗
Filed 2026-09-10T07:54:41Z. Updated 2026-09-10T16:19:14Z. Closed 2026-09-10T16:19:14Z.
CLOSED at capture / #7
erratum: research/prepublication-evidence.md names evidence/reproduce.py, which the mirror does not ship ↗
Filed 2026-09-10T07:40:26Z. Updated 2026-09-24T17:25:19Z. Closed 2026-09-24T17:25:19Z.

Reading guide

A report is not a verdict

The line tracks your place in this explanation.

Reproduce from a public source

The report links the observed behavior and the expected rule. Tool READMEs keep the checks and their limits.

Record the upstream response

Open and closed are GitHub issue states. Closing a report does not establish that every deployed service was fixed.

Recheck before relying

These lists update when a maintainer records the queries again and builds the site. They are not live monitors.