Repository navigation
docs: clarify W1 post-merge verification evidence - #86
Merged
Merged
Conversation
Two low-severity documentation observations from the independent PR #85 review; documentation-only, no evidence artifact touched. 1. Section 4 (zero-movement evidence): the acceptance record implied every adversarial case carries a logs tail alongside the transaction id and slot. In the verbatim JSON artifact all 14 cases carry the rejected transaction id + slot + zero-movement checks, 13 carry a populated logs_tail, and A2's logs_tail is empty — valid evidence, because the A2 rejection occurs at instruction 0 before any program log is emitted. The wording now states exactly that and does not imply missing evidence. 2. Section 3 (toolchain attribution): Anchor CLI 1.2.2 and embedded SBF rustc 1.89.0-dev were listed under the provenance-manifest pins. The manifest pins Solana CLI 3.0.10, SBPF v0, platform-tools v1.54 (archive sha256) and the approved ELF digests — it does not pin the Anchor CLI version or the embedded rustc version as standalone fields. Each value is now attributed to its direct evidence source (check-toolchain.sh output; embedded toolchain information of the SHA-256-pinned platform-tools v1.54 archive). Reported versions unchanged. Verbatim artifacts run-report-2026-11-06T06-49-37-889Z.json and .summary.md remain byte-identical (git blob hashes 7c15595, 1a3a6e3).
graveyieldprotocol
approved these changes
Nov 6, 2026
graveyieldprotocol
left a comment
Collaborator
There was a problem hiding this comment.
Independent review completed — APPROVE. This approval is pinned to head 1a20bfe0a0a6d7d398b7d83d04ddf9fe1f973129.
Verified independently against the live GitHub state and repository contents:
- Scope: exactly one modified file,
docs/w1-post-merge/W1_POST_MERGE_VERIFICATION.md, +16/−5, single commit; the merge-base diff contains nothing else. - Observation 1 (adversarial log-tail wording): checked against
run-report-2026-11-06T06-49-37-889Z.json. All 14 adversarial cases carry a transaction id, a slot,zero_movement.proved = trueand per-check zero-movement evidence; 13 of 14 cases have a populatedlogs_tail; A2's is empty, with the recorded failureInstructionError[0, Custom:2](precompile family) — consistent with the revised instruction-0 / signature-verification-precompile explanation. The new wording does not imply missing transaction evidence and does not weaken the significance of the empty log tail. - Observation 2 (toolchain attribution):
scripts/devnet/provenance.manifest.jsoncontains Solana CLI3.0.10, SBPF v0, platform-toolsv1.54(archive sha256fcc41631…) and the approved Scanner/Vault ELF digests, and carries no Anchor CLI or rustc fields. The revised §3 attributes Anchor CLI1.2.2to thescripts/check-toolchain.shoutput recorded in §3.1, and the embedded SBF rustc1.89.0-devto the sha256-pinned platform-tools v1.54 archive inspection recorded in §3.2. The prior manifest over-claim is corrected. - Historical evidence integrity: the W1.1 JSON report and Markdown summary are byte-identical to
main(git blob equality); no lifecycle/adversarial result, classification, approved hash, or security-boundary statement was altered; nothing was regenerated or replaced. - CI at this exact head: 13 checks passed, 0 failed, 0 pending.
build + fetch + run 7 fork suiteswas skipped by its fail-closed change-detection gate (docs-only diff); the skip is visible by design and is not counted as a pass. CodeQL completed successfully.scripts/doc-links-lint.shandscripts/terminology-lint.shwere re-run locally at this commit: both exit 0. Prettier is out of scope for*.mdper.prettierignore. - Non-blocking, out-of-scope observation:
.github/workflows/fork-harness.ymlline 51 readsbranches: ain](apparently a corruptedbranches: [main]). Pre-existing onmain, empirically inert (pull_request-triggered fork-harness runs fire systematically), and untouched by this PR. Recommend a separate one-line follow-up PR.
Recommendation basis: exact scope, both corrections evidence-supported, historical evidence untouched, CI green with the documented gate-driven skip. PR remains open for maintainer disposition.
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
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.
Summary
Documentation-only PR resolving the two low-severity observations from the independent review of PR #85 (W1.1 post-merge verification evidence). No code, tests, workflows, harnesses, manifests, or approved hashes are touched.
Observation 1 — §4 logs-tail wording
The acceptance record previously stated that every adversarial case carries "the rejected transaction id + slot + logs tail", implying a populated logs tail for all 14 cases. The verbatim JSON artifact shows:
zero_movement.proved = trueand the per-check zero-movement fields;logs_tail;logs_tailis empty — valid evidence, because the A2 rejection occurs at instruction 0 (the signature-verification precompile errors before the program is ever invoked, so no program logs are emitted).§4 now states this precisely, records that logs tails are captured when available, and explicitly notes that an empty logs tail does not indicate missing transaction evidence or a failed adversarial test.
Observation 2 — §3 toolchain citation attribution
§3 previously listed Anchor CLI
1.2.2and embedded SBF rustc1.89.0-devunder the pins "fromscripts/devnet/provenance.manifest.json". The manifest pins Solana CLI3.0.10, SBPF v0, platform-toolsv1.54(archive sha256fcc41631…) and the approved Scanner/Vault ELF digests — it does not pin the Anchor CLI version or the embedded rustc version as standalone fields. §3 now attributes each value to its direct evidence source:1.2.2— thescripts/check-toolchain.shoutput recorded in §3.1;1.89.0-dev— the toolchain information embedded in the SHA-256-pinned platform-tools v1.54 archive (inspection recorded in §3.2).Reported versions are unchanged; no claim is made about manifest fields the manifest does not contain.
Scope verification
git diffvsmaintouches exactly one file:docs/w1-post-merge/W1_POST_MERGE_VERIFICATION.md(16 insertions, 5 deletions).run-report-2026-11-06T06-49-37-889Z.jsonandrun-report-2026-11-06T06-49-37-889Z.summary.mdremain byte-identical tomain(git blob hashes7c155953,1a3a6e3b— unchanged).Documentation checks
scripts/doc-links-lint.sh— 132 relative link targets checked, all resolve (exit 0).scripts/terminology-lint.sh— no forbidden vocabulary detected (exit 0).prettier --check .(3.9.10, repo config) — all files pass, including the edited document (exit 0).