docs: fix "fill or kill type or order" typo in allow_partial doc comments by unarbos · Pull Request #2875 · RaoFoundation/subtensor

unarbos · 2026-07-12T20:21:08Z

Summary

Fixes the recurring "fill or kill type or order" -> "type of order" typo in the allow_partial parameter docs (5 occurrences across dispatches.rs, add_stake.rs, remove_stake.rs).
Doc-comment only; no behavior change. Also serves as the first full canary of the snapshot-cached Runtime Checks pipeline that landed in ci: snapshot-cached runtime checks — split clone shards, required try-runtime snapshots, integrity manifests #2874: this PR touches pallets/, so it exercises the split builds, the required try-runtime snapshot path, and all four clone shards end to end.

Test plan

try-runtime devnet/testnet/mainnet restore v-named snapshots from the freshly dispatched producer run (expected minutes, not a 30-min live scrape)
All four clone-upgrade shards and the Sudo-upgrade gate green

Made with Cursor

…ents Co-authored-by: Cursor <cursoragent@cursor.com>

vercel · 2026-07-12T20:21:11Z

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated (UTC)
subtensor	Ready	Preview, Comment	Jul 13, 2026 3:24pm

github-actions · 2026-07-12T20:22:15Z

🛡️ AI Review — Skeptic (security review)

VERDICT: VULNERABLE

MEDIUM scrutiny — 118-day-old contributor with repository write access and substantial history; no Gittensor association; fix-fill-or-kill-doc-typo → main.

The PR is no longer doc-only: it includes the merged runtime-upgrade signing and CI changes. The previously identified signing trust-boundary issue remains.

Findings

Sev	File	Finding
HIGH	`sdk/python/bittensor/cli/commands/upgrade.py:447`	Require an independently built runtime before signing	inline

Prior-comment reconciliation

a8e52f97: not addressed — The command still proceeds after merely warning that the runtime came from the release rather than a local build.

Conclusion

The change appears legitimate, but the signing command still permits mainnet approval without independently verifying the runtime bytes.

📜 Previous run (superseded)

~~Sev~~	~~File~~	~~Finding~~	Status
~~HIGH~~	~~`sdk/python/bittensor/cli/commands/upgrade.py:447`~~	~~Require an independently built runtime before signing~~	➡️ Carried forward to current findings The command still proceeds after merely warning that the runtime came from the release rather than a local build.

🔍 AI Review — Auditor (domain review)

VERDICT: 👍

UNKNOWN Gittensor association; established high-volume contributor with repository write access.

The changes are non-runtime and require neither tests nor a spec_version bump.

Description discrepancy: the PR body calls this a doc-comment-only change, but the branch also changes Docker-pull retry behavior and adds a scoped deprecation allowance to a migration test. The body should mention those ancillary changes.

Findings

Sev	File	Finding
LOW	`pallets/subtensor/src/macros/dispatches.rs:1393`	Replacement remains grammatically awkward	inline

Prior-comment reconciliation

e930e934: not addressed — The wording remains “fill or kill type of order”; the standard phrasing is “a fill-or-kill order.”

Conclusion

The PR is safe to merge, though the remaining awkward wording should be corrected consistently before or during merge.

📜 Previous run (superseded)

~~Sev~~	~~File~~	~~Finding~~	Status
~~LOW~~	~~`pallets/subtensor/src/macros/dispatches.rs:1393`~~	~~Replacement remains grammatically awkward~~	➡️ Carried forward to current findings The wording remains “fill or kill type of order”; the standard phrasing is “a fill-or-kill order.”

github-actions

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

github-actions · 2026-07-12T20:23:29Z

        /// * `allow_partial`: Allows partial execution of the amount. If set to false, this becomes
-        ///       fill or kill type or order.
+        ///       fill or kill type of order.


[LOW] Replacement remains grammatically awkward

“Fill-or-kill” is the established compound adjective; “fill or kill type of order” remains ungrammatical. Use the clearer wording below consistently across all five changed comments.

Suggested change

/// * `allow_partial`: Allows partial execution of the amount. If set to false, this becomes

/// fill or kill type or order.

/// fill or kill type of order.

/// * `allow_partial`: Allows partial execution of the amount. If set to false, this becomes

/// a fill-or-kill order.

github-actions · 2026-07-12T20:23:37Z

🔄 AI review updated — Skeptic: SAFE Auditor: 👍

…ration test The doc comment typo fix propagates into runtime metadata, so the generated Python calls.py must be regenerated to keep the metadata drift gate green. Also silence clippy::deprecated on the LastTxBlockDelegateTake migration test, which intentionally exercises the deprecated storage aliases. Co-authored-by: Cursor <cursoragent@cursor.com>

github-actions

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

github-actions · 2026-07-13T14:08:25Z

        ///
        /// * `allow_partial`: Allows partial execution of the amount. If set to false, this becomes
-        ///       fill or kill type or order.
+        ///       fill or kill type of order.


[LOW] Replacement remains grammatically awkward

“Fill or kill type of order” is still unnatural. Use the standard compound term “a fill-or-kill order,” and apply the same wording to the other source and generated occurrences in this PR.

Suggested change

/// fill or kill type of order.

/// a fill-or-kill order.

github-actions · 2026-07-13T14:08:35Z

🔄 AI review updated — Skeptic: SAFE Auditor: 👍

…ials With ~120 parallel e2e jobs pulling the same localnet image, GHCR occasionally returns a spurious permission_denied. Retry with backoff instead of failing the shard on the first attempt. Co-authored-by: Cursor <cursoragent@cursor.com>

github-actions

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

github-actions · 2026-07-13T14:22:20Z

        ///
        /// * `allow_partial`: Allows partial execution of the amount. If set to false, this becomes
-        ///       fill or kill type or order.
+        ///       fill or kill type of order.


[LOW] Replacement remains grammatically awkward

“Fill or kill type of order” is still unnatural and does not use the standard term. Use “a fill-or-kill order” here and in every corresponding source/generated occurrence.

Suggested change

/// fill or kill type of order.

/// a fill-or-kill order.

github-actions · 2026-07-13T14:22:30Z

🔄 AI review updated — Skeptic: SAFE Auditor: 👍

…eleases) The release train now publishes each mainnet upgrade proposal as a pre-release v<spec> whose page URL is the single thing to share with signers: it carries the signer instructions and the artifacts (proxy_proxy_blob.hex, subtensor.wasm, digest, upgrade-manifest.json) as stable-URL assets. The mainnet watcher promotes the pre-release to the final release once the upgrade executes instead of creating a new one. On the signer side, a new `btcli upgrade` command group replaces the manual prod-approval.js flow: `pending` discovers proposals purely from chain state (sudo.key() -> SudoUncheckedSetCode proxy -> pending multisig ops), `check` verifies the call data end to end (exact structural re-encoding against live metadata, srtool digest, live sudo key, on-chain call hash — with --wasm pinning it to a self-built runtime), and `sign` runs every check then auto-detects the first/interior/final signer position from chain state and submits the matching approval. `btcli multisig pending` surfaces pending upgrades when the multisig is the chain's sudo key. Co-authored-by: Cursor <cursoragent@cursor.com>

github-actions

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

github-actions · 2026-07-13T15:20:54Z

+    if bundle.wasm_source != "local":
+        app_ctx.output.message(
+            "[yellow]note:[/yellow] the runtime was taken from the release, not a local "
+            "srtool build — anyone reproducing the build should pass --wasm"
+        )


[HIGH] Require an independently built runtime before signing

This warning still allows signing when both the proposal blob and WASM came from the release. Those artifacts share the CI/GitHub trust domain that submitted the on-chain proposal, so their byte equality does not independently establish that the runtime corresponds to reviewed source. If CI, its credentials, or release publication is compromised, the recommended btcli upgrade sign --url ... flow lets triumvirate members approve arbitrary structurally valid runtime code. Preserve the multisig’s independent security boundary by refusing to sign unless --wasm supplies a locally reproduced build.

Suggested change

if bundle.wasm_source != "local":

app_ctx.output.message(

"[yellow]note:[/yellow] the runtime was taken from the release, not a local "

"srtool build — anyone reproducing the build should pass --wasm"

)

if bundle.wasm_source != "local":

app_ctx.output.error(

"refusing to sign without an independently built runtime; pass --wasm "

"with the byte-identical srtool build produced from reviewed source"

)

raise typer.Exit(1)

github-actions · 2026-07-13T15:21:07Z

🔄 AI review updated — Skeptic: VULNERABLE

github-actions

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

github-actions · 2026-07-13T15:24:57Z

+    if bundle.wasm_source != "local":
+        app_ctx.output.message(
+            "[yellow]note:[/yellow] the runtime was taken from the release, not a local "
+            "srtool build — anyone reproducing the build should pass --wasm"
+        )


[HIGH] Require an independently built runtime before signing

This remains only a warning, so every triumvirate signer can approve runtime bytes supplied by the same CI/release trust domain that proposed them. If that domain is compromised, all documented checks can agree on malicious bytes without any signer independently deriving the runtime from source. Refuse signing unless --wasm supplied a locally reproduced build.

Suggested change

if bundle.wasm_source != "local":

app_ctx.output.message(

"[yellow]note:[/yellow] the runtime was taken from the release, not a local "

"srtool build — anyone reproducing the build should pass --wasm"

)

if bundle.wasm_source != "local":

app_ctx.output.error(

"refusing to sign without an independently built runtime",

help="build the tagged source with srtool and pass the result with --wasm",

)

raise typer.Exit(1)

github-actions · 2026-07-13T15:25:06Z

🔄 AI review updated — Skeptic: VULNERABLE

docs: fix "fill or kill type or order" typo in allow_partial doc comm…

f85e422

…ents Co-authored-by: Cursor <cursoragent@cursor.com>

vercel Bot deployed to Preview July 12, 2026 20:22 View deployment

github-actions Bot reviewed Jul 12, 2026

View reviewed changes

UnArbosFive mentioned this pull request Jul 12, 2026

ci: harden and accelerate cached runtime checks #2876

Closed

vercel Bot deployed to Preview July 13, 2026 14:06 View deployment

github-actions Bot reviewed Jul 13, 2026

View reviewed changes

vercel Bot deployed to Preview July 13, 2026 14:20 View deployment

github-actions Bot reviewed Jul 13, 2026

View reviewed changes

UnArbosFive added 15 commits July 13, 2026 16:47

ci: clean up snapshot cache controls

098dd84

ci: move scheduled smoke checks to hosted runners

b2311f2

ci: increase snapshot RPC concurrency

b2ebcf1

ci: speed up cached snapshot downloads

a51cec7

revert: retain faster snapshot download path

033e48b

ci: benchmark parallel snapshot downloads

62868fe

ci: benchmark mainnet restore across runner pools

6af747f

ci: parallelize cached snapshot downloads

435766f

ci: benchmark try-runtime on turbo16

f47e986

ci: run try-runtime on turbo16

df09700

ci: run try-runtime on GitHub-hosted runners

f9ffd00

ci: harden cached snapshot workflows

c0bf45d

ci: consolidate and accelerate clone regressions

5b0f2eb

ci: trust base revision for runtime path filtering

2fbd814

ci: test snapshot producer on hosted runners

f70fa9e

UnArbosFive added 18 commits July 13, 2026 16:47

ci: retire dedicated try-runtime runners

650ccd4

ci: drop hosted-only snapshot telemetry

35aee69

ci: cache SDK core builds in clone checks

9b2e973

ci: drop ineffective SDK compiler cache

b7e089b

ci: respect clone RPC page limit

1687152

ci: simplify clone runtime orchestration

0fffc88

ci: benchmark single-node e2e finality

1a830d1

ci: tune single-node e2e block waits

c361271

test: synchronize accelerated evm nonces

7a05c04

ci: validate balanced shield shards

7ab7d2d

ci: finalize accelerated typescript e2e suite

ad25548

test: wait for frontier after substrate transfer

089f96c

test: poll frontier balance after substrate transfer

7525fc9

test: bypass ethers balance cache

e037eda

ci: restore six-hour smoke cadence

d890617

test: bypass cached EVM nonce reads

e45c8a7

ci: shorten typescript e2e critical path

6318e45

ci: keep faster e2e linker path

c83cdba

UnArbosFive mentioned this pull request Jul 13, 2026

ci: accelerate runtime, clone, and TypeScript E2E checks #2882

Merged

UnArbosFive and others added 3 commits July 13, 2026 17:01

ci: centralize e2e setup and validate shield shards

d823013

Merge branch 'pr-2882' into fix-fill-or-kill-doc-typo

c87ac28

vercel Bot deployed to Preview July 13, 2026 15:20 View deployment

github-actions Bot reviewed Jul 13, 2026

View reviewed changes

vercel Bot deployed to Preview July 13, 2026 15:24 View deployment

github-actions Bot reviewed Jul 13, 2026

View reviewed changes

IntiTechnologies merged commit 7b41d53 into main Jul 13, 2026
175 of 177 checks passed

Uh oh!

Conversation

unarbos commented Jul 12, 2026

Summary

Test plan

Uh oh!

vercel Bot commented Jul 12, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

github-actions Bot commented Jul 12, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

🛡️ AI Review — Skeptic (security review)

Findings

Prior-comment reconciliation

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 Jul 12, 2026

Choose a reason for hiding this comment

Uh oh!

github-actions Bot commented Jul 12, 2026

Uh oh!

github-actions Bot left a comment

Choose a reason for hiding this comment

Uh oh!

github-actions Bot Jul 13, 2026

Choose a reason for hiding this comment

Uh oh!

github-actions Bot commented Jul 13, 2026

Uh oh!

github-actions Bot left a comment

Choose a reason for hiding this comment

Uh oh!

github-actions Bot Jul 13, 2026

Choose a reason for hiding this comment

Uh oh!

github-actions Bot commented Jul 13, 2026

Uh oh!

github-actions Bot left a comment

Choose a reason for hiding this comment

Uh oh!

github-actions Bot Jul 13, 2026

Choose a reason for hiding this comment

Uh oh!

github-actions Bot commented Jul 13, 2026

Uh oh!

github-actions Bot left a comment

Choose a reason for hiding this comment

Uh oh!

github-actions Bot Jul 13, 2026

Choose a reason for hiding this comment

Uh oh!

github-actions Bot commented Jul 13, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

vercel Bot commented Jul 12, 2026 •

edited

Loading

github-actions Bot commented Jul 12, 2026 •

edited

Loading