From f3a31df5958fe3b6aa6db90456ab1e650fa957ba Mon Sep 17 00:00:00 2001 From: Nick Date: Thu, 12 Nov 2026 13:14:29 +0000 Subject: [PATCH] release: v0.5.34 Two merged PRs since 0.5.33, no milestone acceptance: M14 slice 2 codex half (#188) + audit claim harvesting for pnpm/yarn and vitest/jest (#190, fixes #189). Co-authored-by: Nick <209239467+surlymochan@users.noreply.github.com> --- CHANGELOG.md | 10 ++++++++++ RELEASE_NOTES.md | 10 ++++++++++ docs/launch-dossier.md | 2 +- package-lock.json | 4 ++-- package.json | 2 +- 5 files changed, 24 insertions(+), 4 deletions(-) diff --git a/CHANGELOG.md b/CHANGELOG.md index f1602f9..0dd9128 100644 --- a/CHANGELOG.md +++ b/CHANGELOG.md @@ -3,6 +3,16 @@ All notable changes to this project are documented here. Detailed notes live in [RELEASE_NOTES.md](RELEASE_NOTES.md). +## [0.5.34] - 2026-11-12 + +### Added +- **M14 slice 2, codex half (PR #188, merged 2026-11-12T09:29:33Z, squash 86215407; evidence issue #187)**: replay of 3 real codex CLI streams recorded on gocc (codex-cli 0.166.0, read-only prompts, stdin from /dev/null) — normal `ls` (9 lines, 2 shell calls), normal `cat sample.txt` (7 lines, 1 shell call), killed `sleep 45` (4 lines, SIGTERM once item.started observed +5 s then SIGKILL, exit 143, no turn.completed). Replay (sink + checkPairing): pairing violations 0 on all three runs; the killed run's single call ends abandoned, 0 calls left running. Acceptance-2 terminal-state views render the same sequence. Tests: `test/agent-stream.test.js` 33 pass; full `npm test` 295/296, the 1 failure `test/i6d-acceptance-demo.test.js` reproduced identically on unmodified main (8081e71) — pre-existing environment failure. + +### Fixed +- **Audit claim harvesting for pnpm/yarn and vitest/jest (PR #190, merged 2026-11-12T11:08:53Z, squash acef97bf; fixes issue #189 by external contributor MAOLVLHL)**: `xflow audit` no longer harvests pnpm runs as npm runs and now picks up yarn/vitest/jest claims. Second-party verification on the PR head (6b894521): module-level checks 12 pass; four suites 126 tests, 123 pass, the 3 failures same names and same cause as the clean-main baseline (pre-existing I6d pytest cases, `pytest` not on PATH in this VM), zero regression; `bin/xflow.js` mode 100755 kept; PR CI ledger + verify green; merged-main CI run 43206616380 success. + +Honest note, stated plainly: milestone M14 is **not** accepted in this release — slice 2's claude half still awaits a real `claude` stream-json recording made in the user's own terminal (issue #187 OPEN, codex half only). No milestone claim and no bench numbers in this release: the issue #5 ledger is settled as a conclusion on the MiniMax full-matrix recheck (135/135 complete, 2026-11-10) — P2-1 passed, P1-2 / P1-3 / P2-3 did not pass. + ## [0.5.33] - 2026-11-12 ### Fixed diff --git a/RELEASE_NOTES.md b/RELEASE_NOTES.md index 82b23b8..d075752 100644 --- a/RELEASE_NOTES.md +++ b/RELEASE_NOTES.md @@ -1,5 +1,15 @@ # Release Notes +## 0.5.34 - Two merged PRs, no milestone acceptance + +Patch release. Two merged PRs since 0.5.33, verified end to end in CI and locally — M14 is not accepted yet, so this release carries no milestone claim. + +**M14 slice 2, codex half (PR #188, merged 2026-11-12T09:29:33Z, squash 86215407; evidence issue #187).** Replay of 3 real codex CLI streams recorded on gocc (codex-cli 0.166.0): normal `ls` (9 lines, 2 shell calls), normal `cat sample.txt` (7 lines, 1 shell call), killed `sleep 45` (4 lines, exit 143, no turn.completed). Pairing violations 0 on all three replays; the killed run's single call ends abandoned, 0 calls left running. Tests: `test/agent-stream.test.js` 33 pass; full `npm test` 295/296, the 1 failure `test/i6d-acceptance-demo.test.js` reproduced identically on unmodified main (8081e71) — pre-existing environment failure. + +**Audit claim harvesting for pnpm/yarn and vitest/jest (PR #190, merged 2026-11-12T11:08:53Z, squash acef97bf; fixes issue #189 by external contributor MAOLVLHL).** `xflow audit` no longer harvests pnpm runs as npm runs and now picks up yarn/vitest/jest claims. Second-party verification on the PR head: module-level checks 12 pass; four suites 126 tests, 123 pass, the 3 failures same names and same cause as the clean-main baseline (pre-existing I6d pytest cases, `pytest` not on PATH in this VM), zero regression; `bin/xflow.js` mode 100755 kept; PR CI ledger + verify green; merged-main CI run 43206616380 success. + +Honest note, stated plainly: milestone M14 is **not** accepted in this release — slice 2's claude half still awaits a real `claude` stream-json recording made in the user's own terminal (issue #187 OPEN, codex half only). No bench numbers in this release: the issue #5 ledger is settled as a conclusion on the MiniMax full-matrix recheck (135/135 complete, 2026-11-10) — P2-1 passed, P1-2 / P1-3 / P2-3 did not pass. + ## 0.5.33 - Milestone M13 "声称抽取 A/B 口径修复" accepted Patch release. One merged PR since 0.5.32, verified end to end in CI and locally — and the milestone it completes is now accepted. diff --git a/docs/launch-dossier.md b/docs/launch-dossier.md index 10bf3ff..f6a207a 100644 --- a/docs/launch-dossier.md +++ b/docs/launch-dossier.md @@ -117,7 +117,7 @@ Evidence: - Adoption records: 1; status: pass - Goal audit: pass; goal: Keep evalid credible as an open-source agent delivery runtime: repo-owned workflow contracts, executable verification gates, durable project context, and honest benchmark evidence. - Package evidence: missing; source: registry_json -- Package name/version: @surlymochan/evalid@0.5.33 +- Package name/version: @surlymochan/evalid@0.5.34 ## Passing Evidence diff --git a/package-lock.json b/package-lock.json index d152537..10bb72d 100644 --- a/package-lock.json +++ b/package-lock.json @@ -1,12 +1,12 @@ { "name": "@surlymochan/evalid", - "version": "0.5.33", + "version": "0.5.34", "lockfileVersion": 3, "requires": true, "packages": { "": { "name": "@surlymochan/evalid", - "version": "0.5.33", + "version": "0.5.34", "license": "MIT", "dependencies": { "ajv": "^8.12.0", diff --git a/package.json b/package.json index 47602b9..83b922b 100644 --- a/package.json +++ b/package.json @@ -1,6 +1,6 @@ { "name": "@surlymochan/evalid", - "version": "0.5.33", + "version": "0.5.34", "description": "Claimed != verified: a runtime verification layer for coding agents. It seals each test run and replays it from a clean snapshot, with a zero-judge gate pipeline (spec clarification, code review, TDD proof) and the xflow CLI.", "keywords": [ "ai-agents",