Skip to content
Closed
Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
31 commits
Select commit Hold shift + click to select a range
8320dda
refactor(review): Extract npm toolchain adapter
wenshao Aug 2, 2026
50a91f4
feat(review): Add Maven multi-module verification
wenshao Aug 2, 2026
ad4bab5
test(review): pin adapter delegation, exec fallback, single-root guard
wenshao Aug 2, 2026
4c1b3bb
Merge branch 'refactor/review-toolchain-adapter' of https://github.co…
qwen-code-dev-bot Aug 2, 2026
187a272
test(review): align Agent 7 brief assertion with build-test command (…
qwen-code-dev-bot Aug 2, 2026
ab61037
fix(review): Tighten Maven verification evidence
wenshao Aug 2, 2026
5fa39d6
Merge remote-tracking branch 'wenshao/refactor/review-toolchain-adapt…
wenshao Aug 2, 2026
2f9a3e1
Merge branch 'main' into refactor/review-toolchain-adapter
qwen-code-dev-bot Aug 3, 2026
f7602ea
fix(review): Reject inactive nested Maven projects
wenshao Aug 3, 2026
f308f0b
fix(review): Tighten Maven evidence, scope, and base-tree gate (#8394)
qwen-code-dev-bot Aug 3, 2026
7f0f42c
fix(review): Address Maven verification review feedback (#8394)
qwen-code-dev-bot Aug 3, 2026
856666f
fix(review): Clamp Maven test counts and skip doc-only module runs (#…
qwen-code-dev-bot Aug 3, 2026
117dbe7
fix(review): Harden Maven classification and Test Plan rulings (#8394)
qwen-code-dev-bot Aug 3, 2026
4304843
fix(review): Close round-3 Maven verification and Test Plan gaps (#8394)
qwen-code-dev-bot Aug 3, 2026
e425e0e
fix(review): Keep Maven source failures out of the infra carve-out (#…
qwen-code-dev-bot Aug 4, 2026
69dc0ee
Merge branch 'main' into refactor/review-toolchain-adapter
wenshao Aug 4, 2026
d25cad5
Merge branch 'main' into refactor/review-toolchain-adapter
qwen-code-dev-bot Aug 4, 2026
923ee6e
fix(review): Close round-4 Maven evidence and Test Plan gaps (#8394)
qwen-code-dev-bot Aug 4, 2026
c403ccc
fix(review): Close round-5 toolchain settlement and classification ga…
qwen-code-dev-bot Aug 5, 2026
912641f
fix(review): Close round-6 diagnostic, deadline, and hardening gaps (…
qwen-code-dev-bot Aug 5, 2026
b6e2bb5
Merge branch 'main' into refactor/review-toolchain-adapter
qwen-code-dev-bot Aug 5, 2026
fc6092f
Merge main into review-toolchain adapter; close round-6 findings (#8394)
qwen-code-dev-bot Aug 5, 2026
c97cf65
Merge branch 'main' into refactor/review-toolchain-adapter
qwen-code-dev-bot Aug 6, 2026
96101d3
Merge branch 'main' into refactor/review-toolchain-adapter
qwen-code-dev-bot Aug 6, 2026
fc6bcb1
fix(review): Close round-7 budget, parser, and settlement gaps (#8394)
qwen-code-dev-bot Aug 6, 2026
4fbc7f3
Merge branch 'main' into refactor/review-toolchain-adapter
qwen-code-dev-bot Aug 6, 2026
c905cf3
Merge branch 'main' into refactor/review-toolchain-adapter
qwen-code-dev-bot Aug 7, 2026
8a27463
fix(review): Close round-8 Maven routing, settlement, and evidence ga…
qwen-code-dev-bot Aug 7, 2026
db9ffc7
fix(review): Close round-9 Maven evidence-gating and deadline defects…
qwen-code-dev-bot Aug 7, 2026
2be30c5
fix(review): Close round-10 Maven adapter bounds, settlement, and evi…
qwen-code-dev-bot Aug 7, 2026
28c13cd
fix(review): Close round-11 Maven evidence, inheritance, and settleme…
qwen-code-dev-bot Aug 8, 2026
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
488 changes: 488 additions & 0 deletions docs/design/review-toolchain-adapters.md

Large diffs are not rendered by default.

8 changes: 7 additions & 1 deletion packages/cli/src/commands/review/agent-prompt.test.ts
Original file line number Diff line number Diff line change
Expand Up @@ -2081,8 +2081,14 @@ describe('buildRoleBrief — every agent, not just the territory ones', () => {
// send the reader to fix a prompt that is correct.
const p = buildRoleBrief(PR_PLAN, '7');
expect(p).not.toContain(PLAN.diffPathAbsolute);
expect(p).toContain('npm run build');
expect(p).toContain('`toolchain: "maven"`');
expect(p).toContain('Do not run `test-delta` for Maven in this release');
Comment on lines +2084 to +2085

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

[Suggestion] R5-8: The assertion this pair replaced (toContain('npm run build')) was the only pin of Agent 7's no-substitution rule; the rewritten rule "Do not substitute hand-written npm or Maven commands." (lib/agent-briefs.ts:448 — verified verbatim, exactly one occurrence repo-wide) is pinned by no test. Deleting that sentence ships green, and it is the only steering against hand-run full builds — the pattern the same paragraph records as timing out 71 times and verifying nothing. The house's own practice pins this class of brief prose (the verify brief's "do not just trace it — run it"). — Failure scenario: a future edit deletes the sentence → the entire suite stays green → Agent 7 can silently regress to hand-run full builds.

    expect(p).toContain('Do **not** substitute hand-written npm or Maven commands');
中文说明

这对断言替换掉的 toContain('npm run build') 是 Agent 7 禁止手写命令规则的唯一钉住;新规则"Do not substitute hand-written npm or Maven commands."(lib/agent-briefs.ts:448,已核实原文全仓仅一处)没有任何测试钉住。删除该句整套测试仍绿,而它是唯一阻止手写全量构建的约束——同一段落记载该模式曾超时 71 次且什么都没验证。仓库自身惯例就是钉住这类简报语句(verify 简报的 "do not just trace it — run it")。失败场景:未来编辑删除该句 → 全套测试保持绿色 → Agent 7 可静默退回手写全量构建。

— qwen3.8-max via Qwen Code /review (v0.21.5)

expect(p).toContain('Source: [build]');
// The only steering against hand-run full builds — the pattern the same
// paragraph records as timing out 71 times and verifying nothing.
expect(p).toContain(
'Do **not** substitute hand-written npm or Maven commands',
);
});

it('pins Agent 7 to the PR worktree and hands it the test-efficacy probe', () => {
Expand Down
Loading
Loading