Subnet Protocol Alpha Accounting by JohnReedV · Pull Request #2645 · opentensor/subtensor

JohnReedV · 2026-05-07T14:09:10Z

Description

This PR adds protocol-owned alpha accounting for subnet chain buys and includes that alpha in subnet deregistration settlement.

Stops recycling alpha bought during protocol TAO buys
Adds per-subnet cached protocol alpha accounting
Includes protocol alpha-in and cached protocol alpha in deregistration pro-rata settlement
Clears cached protocol alpha when a subnet is dissolved
Updates tests for the new deregistration settlement behavior

github-actions · 2026-05-26T14:56:21Z

🛡️ AI Review — Skeptic (security review)

VERDICT: SAFE

Baseline scrutiny: established Opentensor contributor with repo write permission; branch chain-buy-cache -> devnet-ready; no trusted Gittensor allowlist hit.

Static review only. The PR does not modify .github/, dependency manifests, lockfiles, or build scripts. I reviewed the protocol-alpha cache, deregistration settlement changes, storage cleanup, arithmetic, generated weight deltas, and contributor/commit signals for malicious behavior, panic surface, origin bypasses, supply-chain risk, hidden activation conditions, and fund-flow security regressions; I did not find a Skeptic-blocking issue.

The known dissolve weight-accounting mismatch remains an Auditor-tracked domain finding, not a Skeptic security finding here because the affected calls are root-only.

Findings

No findings.

Conclusion

No malicious behavior or security vulnerability was found in the reviewed diff.

🔍 AI Review — Auditor (domain review)

VERDICT: 👍

Established opentensor contributor with repo write permission and long subtensor history; Gittensor association UNKNOWN from trusted allowlists.

PR body is substantive and matches the implementation. I did not find a duplicate-work concern among the overlapping open PRs; the overlaps appear to be adjacent subtensor/runtime churn rather than competing implementations.

git diff --check passed. I did not run targeted Rust tests because the carried-forward issue is a static weight-accounting mismatch. Spec-version auto-fix was not applied because the devnet RPC host could not be resolved from this sandbox, so I could not safely compare local spec_version: 412 against live devnet.

Findings

Sev	File	Finding
MEDIUM	`pallets/subtensor/src/staking/remove_stake.rs:477`	Update dissolve weights for protocol-alpha storage	inline

Prior-comment reconciliation

1969a08a: not addressed — SubnetProtocolAlpha is still absent from the hard-coded dissolve_network / root_dissolve_network weight accounting; the weights remain reads(6) / writes(31).

Conclusion

The protocol-alpha accounting change is consistent with the PR description and has focused tests for coinbase caching and deregistration settlement. The carried-forward weight-accounting issue remains medium severity because the affected dissolve calls are root-only, so I am not blocking merge on it.

📜 Previous run (superseded)

~~Sev~~	~~File~~	~~Finding~~	Status
~~MEDIUM~~	~~`pallets/subtensor/src/staking/remove_stake.rs:477`~~	~~Update dissolve weights for protocol-alpha storage~~	➡️ Carried forward to current findings `SubnetProtocolAlpha` is still absent from the hard-coded `dissolve_network` / `root_dissolve_network` weight accounting; the weights remain `reads(6)` / `writes(31)`.

github-actions

AI review — see the sticky summary comment for the verdict and the inline comments below for specific findings.

github-actions · 2026-05-26T15:01:01Z

+        let protocol_alpha_value_u128: u128 = SubnetAlphaIn::<T>::get(netuid)
+            .saturating_add(SubnetProtocolAlpha::<T>::get(netuid))
+            .to_u64() as u128;


[MEDIUM] Update dissolve weights for protocol-alpha storage

This adds a new SubnetProtocolAlpha read to the dissolve path, and the same function now also removes that map later in the cleanup. The dispatch weights for both dissolve_network and root_dissolve_network in pallets/subtensor/src/macros/dispatches.rs still charge the old reads(6) / writes(31), so the runtime undercharges this path by at least the new protocol-alpha read/write. Please update those annotations or regenerate the relevant weight so the storage proof and DB ops match the new dissolve accounting.

github-actions · 2026-05-26T15:01:11Z

🔄 AI review updated — Skeptic: SAFE Auditor: 👍

github-actions

AI review — see the sticky summary comment for the verdict and the inline comments below for specific findings.

github-actions · 2026-05-31T15:40:53Z

+        let protocol_alpha_value_u128: u128 = SubnetAlphaIn::<T>::get(netuid)
+            .saturating_add(SubnetProtocolAlpha::<T>::get(netuid))


[MEDIUM] Update dissolve weights for protocol-alpha storage

This adds a SubnetProtocolAlpha read to the dissolve cleanup path, and the same function also removes SubnetProtocolAlpha later. The static weights for dissolve_network and root_dissolve_network in pallets/subtensor/src/macros/dispatches.rs still use reads(6) / writes(31) and do not account for this new storage item. Update those weights at minimum by +1 read/+1 write, or regenerate/replace the hard-coded dissolve weights with a benchmark that exercises SubnetProtocolAlpha.

github-actions · 2026-05-31T15:41:01Z

🔄 AI review updated — Skeptic: SAFE Auditor: 👍

github-actions

AI review — see the sticky summary comment for the verdict and the inline comments below for specific findings.

github-actions · 2026-05-31T15:46:24Z

+        let protocol_alpha_value_u128: u128 = SubnetAlphaIn::<T>::get(netuid)
+            .saturating_add(SubnetProtocolAlpha::<T>::get(netuid))


[MEDIUM] Update dissolve weights for protocol-alpha storage

destroy_alpha_in_out_stakes now reads SubnetProtocolAlpha here and later removes it, but the dissolve_network / root_dissolve_network dispatch weights are still hard-coded at reads(6) / writes(31). Please add the new read/write to those weights, or re-benchmark and wire the dissolve weight through generated weights, so the root dissolve path charges for the extra storage access.

github-actions · 2026-05-31T15:46:33Z

🔄 AI review updated — Skeptic: SAFE Auditor: 👍

github-actions

AI review — see the sticky summary comment for the verdict and the inline comments below for specific findings.

github-actions · 2026-05-31T16:22:09Z

+        let protocol_alpha_value_u128: u128 = SubnetAlphaIn::<T>::get(netuid)
+            .saturating_add(SubnetProtocolAlpha::<T>::get(netuid))


[MEDIUM] Update dissolve weights for protocol-alpha storage

do_dissolve_network now reads SubnetProtocolAlpha and later removes it, but the hard-coded weights on both dissolve_network and root_dissolve_network still charge only reads(6) / writes(31). Please account for the extra storage read and write on those dispatch weights. This is root-only, so I am keeping it medium severity rather than blocking as a user-callable weight undercharge.

github-actions · 2026-05-31T16:22:17Z

🔄 AI review updated — Skeptic: SAFE Auditor: 👍

github-actions

AI review — see the sticky summary comment for the verdict and the inline comments below for specific findings.

github-actions · 2026-05-31T16:52:25Z

        let mut stakers: Vec<(T::AccountId, T::AccountId, u128)> = Vec::new();
-        let mut total_alpha_value_u128: u128 = 0;
+        let protocol_alpha_value_u128: u128 = SubnetAlphaIn::<T>::get(netuid)
+            .saturating_add(SubnetProtocolAlpha::<T>::get(netuid))


[MEDIUM] Update dissolve weights for protocol-alpha storage

This adds protocol-alpha storage access during subnet dissolve, but the call weights for dissolve_network / root_dissolve_network in pallets/subtensor/src/macros/dispatches.rs are still hard-coded at reads(6) / writes(31). The new flow reads SubnetAlphaIn and SubnetProtocolAlpha for the pro-rata denominator and removes SubnetProtocolAlpha during cleanup, so the hard-coded weights under-account the storage work. Update both dissolve call weights to include the added reads/writes, or wire them through a benchmarked WeightInfo entry instead of leaving the stale constants.

github-actions · 2026-05-31T16:52:35Z

🔄 AI review updated — Skeptic: SAFE Auditor: 👍

impl SubnetProtocolAlpha

bbb22e8

JohnReedV added the skip-cargo-audit This PR fails cargo audit but needs to be merged anyway label May 7, 2026

JohnReedV added 4 commits May 7, 2026 07:37

bump spec

8344515

bump spec again

311fcc5

Merge branch 'devnet-ready' into chain-buy-cache

9b91b4b

Merge branch 'devnet-ready' into chain-buy-cache

ea9ed53

gztensor requested changes May 15, 2026

View reviewed changes

Comment thread pallets/subtensor/src/coinbase/run_coinbase.rs

JohnReedV added the apply-benchmark-patch label May 15, 2026

auto-update benchmark weights

695a6d8

github-actions Bot removed the apply-benchmark-patch label May 15, 2026

Merge branch 'devnet-ready' into chain-buy-cache

bb33770

JohnReedV added the apply-benchmark-patch label May 15, 2026

auto-update benchmark weights

de0e9f2

github-actions Bot removed the apply-benchmark-patch label May 15, 2026

JohnReedV force-pushed the chain-buy-cache branch from 7a2985d to de0e9f2 Compare May 15, 2026 18:01

github-actions Bot mentioned this pull request May 21, 2026

Ai reviewer #2668

Merged

4 tasks

Merge branch 'devnet-ready' into chain-buy-cache

20b8401

gztensor previously approved these changes May 26, 2026

View reviewed changes

github-actions Bot reviewed May 26, 2026

View reviewed changes

JohnReedV added 3 commits May 31, 2026 08:34

trigger CI

9fef2d6

Merge branch 'devnet-ready' into chain-buy-cache

c4f9623

conflict

646b742

JohnReedV dismissed gztensor’s stale review via 646b742 May 31, 2026 15:40

github-actions Bot reviewed May 31, 2026

View reviewed changes

fix wieghts

32df63c

github-actions Bot reviewed May 31, 2026

View reviewed changes

JohnReedV added the apply-benchmark-patch label May 31, 2026

auto-update benchmark weights

255390f

github-actions Bot removed the apply-benchmark-patch label May 31, 2026

trigger CI

66cee25

github-actions Bot reviewed May 31, 2026

View reviewed changes

github-actions Bot mentioned this pull request May 31, 2026

Record miner burn as TAO outflow in coinbase #2692

Open

3 tasks

sam0x17 approved these changes Jun 1, 2026

View reviewed changes

		let protocol_alpha_value_u128: u128 = SubnetAlphaIn::<T>::get(netuid)
		.saturating_add(SubnetProtocolAlpha::<T>::get(netuid))

Conversation

JohnReedV commented May 7, 2026

Description

Uh oh!

Uh oh!

github-actions Bot commented May 26, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

🛡️ AI Review — Skeptic (security review)

Findings

Conclusion

🔍 AI Review — Auditor (domain review)

Findings

Prior-comment reconciliation

Conclusion

Uh oh!

github-actions Bot left a comment

Choose a reason for hiding this comment

Uh oh!

github-actions Bot May 26, 2026

Choose a reason for hiding this comment

Uh oh!

github-actions Bot commented May 26, 2026

Uh oh!

github-actions Bot left a comment

Choose a reason for hiding this comment

Uh oh!

github-actions Bot May 31, 2026

Choose a reason for hiding this comment

Uh oh!

github-actions Bot commented May 31, 2026

Uh oh!

github-actions Bot left a comment

Choose a reason for hiding this comment

Uh oh!

github-actions Bot May 31, 2026

Choose a reason for hiding this comment

Uh oh!

github-actions Bot commented May 31, 2026

Uh oh!

github-actions Bot left a comment

Choose a reason for hiding this comment

Uh oh!

github-actions Bot May 31, 2026

Choose a reason for hiding this comment

Uh oh!

github-actions Bot commented May 31, 2026

Uh oh!

github-actions Bot left a comment

Choose a reason for hiding this comment

Uh oh!

github-actions Bot May 31, 2026

Choose a reason for hiding this comment

Uh oh!

github-actions Bot commented May 31, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

github-actions Bot commented May 26, 2026 •

edited

Loading