@tea-agent/loop-agent 0.23.1 → 0.24.1

package/CHANGELOG.md

@@ -2,6 +2,67 @@
 
 ## [Unreleased]
 
+### 改进
+
+- 夜间自动发布固定为 patch 通道（当前 `0.x.y` 上 +0.0.1，例如 `0.24.1 -> 0.24.2`），不再因 `feat:` 自动跳到 `0.x.0`。
+- 新增人工主版本工作流 `manual-minor-release.yml`，仅在显式触发时强制 bump minor，产出 `v0.x.0`。
+- 发布通道可将文档版本钉死/漂移降为告警（`LOOP_AGENT_DOC_VERSION_REFS=warn`），避免 nightly/prepublish 被安装文档旧精确串误阻塞；结构性文档治理仍 fail-closed。
+
+### 修复
+
+- 发布流程不再被合法的旧版本引用误阻塞：文档 catalog 现可用 `current`/`floating`/`minimum`/`range`/`historical` 五类语义声明版本引用，审查时基线等历史证据保留当时真实精确版本并通过检查，仅当前安装入口的精确钉死仍判为漂移。
+- Nightly Release 现在在版本号与 CHANGELOG 准备完成之后、push/打 tag/发布 npm/创建 GitHub Release 之前运行最终 release tree 门禁（测试、构建、仓库治理、`npm pack` 干跑），避免在未通过最终门禁的发布树产生远端副作用。
+- Nightly Release 的 commit/tag、npm publish、GitHub Release 三步现支持幂等恢复：重跑时探测目标对象是否存在，一致则跳过、冲突则 fail closed（不 force-push、不重写 tag、不重复发布），可补齐半发布状态。
+- 文档版本治理的 `minimum`/`range` 声明现会校验文档正文实际出现的精确版本串与声明门槛/范围一致；仅元数据通过而正文钉死低于门槛或不满足范围的版本时仍判失败，避免安装入口借 minimum/range 声明悄悄钉死旧版本。
+
+## [0.24.1] - 2026-07-27
+
+### 重点更新
+
+- 后端测试 HTML 报告全面升级，采用更清晰的可折叠用例卡片设计，直观展示脱敏的请求与响应详情。
+- 彻底修复了 Windows 环境下因 cmd.exe 与 bash 嵌套引号转义不当，导致前端与初始化测试固件执行失败的问题。
+- 前端实现 DAG 大幅优化，复用现有 run-owned canonical artifacts，有效精简了常规全绿路径的直接依赖节点数量。
+
+### 新增
+
+- 后端测试第 7 节点现使用 pytest-html 生成自包含报告，并新增解析器从其 JSON island 提取每条用例的结果与日志。
+- 后端测试生成的 pytest 文件名现可与 Markdown 模块建立一对一清晰映射。
+
+### 改进
+
+- 后端测试 HTML 报告对齐 R04 模板样式：改用全 inline style 设计，包含渐变 banner、可折叠用例卡片、状态胶囊以及请求参数与响应结果分块展示。
+- 前端 canonical implementation contract 允许纯逻辑修改使用空的 UI state，不再强制虚构，并会确定性移除不适用状态下的空输出。
+- 前端实现 DAG 的多个节点（如 frontend-implement-pi、frontend-review-pi 等）直接依赖得到收敛，不再重复读取完整任务源或上游冗余文本。
+
+### 修复
+
+- 修复后端测试 HTML 报告中请求参数显示为空的问题：渲染器现已兼容 payload= 键或 JSON request_record 格式的请求摘要。
+- 修复后端测试报告无法获取请求/响应日志的问题：兼容 HTTP_REQUEST 与 HTTP REQUEST 两种日志前缀，并在解析前正确解码 HTML 实体。
+- 修复 shell-executor 在 Windows 环境下因反斜杠路径导致 runDir 损坏的问题，确保路径注入稳定。
+- 修复前端 worktree baseline 解析 git status 时因提前 trim 导致未暂存文件路径丢失首字符的解析缺陷。
+- 统一 DAG 任务源路径生成与运行时解析基准，解决 freshness、rerun 与预览共享不一致的问题。
+- 修复前端项目在需求未显式提及 UI 时被误判保留为标准任务的问题，提升了基于工程证据与交互语义的任务类型推断准确度。
+- 修复前端 lint baseline 在 Windows 下因路径解析逻辑缺陷，拒绝相对路径输出导致评估误报为不可用的问题。
+- 修复多个测试固件在 Windows 下因嵌套 Shell 引号剥离导致命令执行输出为空或报错的问题（涉及 composite-shell 与 init-command）。
+
+## [0.24.0] - 2026-07-27
+
+### 新增
+
+- align chat workspace with pi web UX
+- operator chat workspace UX + async runtime race guards
+- operator chat M3-M7 (draft interview, human-gate, input context, artifacts, branching, landing)
+- pytest-html report + filename mapping + Windows path fix (#71)
+
+### 改进
+
+- style(console): 收紧 Operator Chat 工作台密度与侧栏体验
+- style(console): prettier-format Operator Chat + keep UI-contract text intact
+
+### 修复
+
+- operator chat M2 closure tracer + code review followups
+
 ## [0.23.1] - 2026-07-26
 
 ### 重点更新
@@ -69,7 +130,6 @@
 - Console Phase 4 Chat P1 审查修复：`createSessionFromServices` 显式传 `tools:[bash,read,grep,find,ls]`（SDK 默认 builtin 不含 grep/find/ls，否则 setActiveToolsByName 静默 no-op）；schema↔dispatcher 契约对齐（contractValidate/Diff 用 draftJson、workerTaskRetry 补 featureId、动态发现 fail-closed）；server.close() 先 chatRuntime.disposeAll()；resource-loader/system prompt 统一到 widening（hasBash:true）
 - CI 治理：7 个 versionConsumer 文档版本漂移 0.22.0→0.23.0；check-task-pool-root.sh allowlist 增补 codify-constraints.md
 
-
 ## [0.22.0] - 2026-07-24
 
 ### 重点更新

package/README.md

@@ -100,7 +100,7 @@ agent-worker console doctor --repo .
 
 ```
 
-访问 Console <http://127.0.0.1:8790/>；顶部「检视」以及 `/inspect/` 提供原 Observe 的只读 Dashboard、DAG、时间线与 artifacts。需要保留的旧书签或脚本仍可在兼容窗口使用 Observe `:8787`。openCode 等主会话仍是 **Compatibility / Operator Assist**，不得升级为与 Official 同等保证。
+访问 Console <http://127.0.0.1:8790/>；**Operator Chat Official** 是 Console 内的主引导交互。构建时设置 `VITE_OPERATOR_CHAT_DEFAULT_LANDING=chat` 可将 Chat 作为默认 landing；flag 未开启时仍默认进入 Tasks，旧 Task Ops tab 始终保留。顶部「检视」以及 `/inspect/` 提供只读 Dashboard、DAG、时间线与 artifacts；`loop-agent …` / `agent-worker …` CLI direct path 仍是受治理 fallback。当前状态为 **code ready, dogfood pending/unmet**，不宣称已替代 openCode 或其他外部主会话。
 
 ## 核心边界
 

package/dist/commands/init.js

@@ -2049,8 +2049,8 @@ function buildTargetFeatureWorkflow(input) {
         "",
         "## Specialized Task Kinds",
         "",
-        "- standard tasks are classified conservatively from the task title, `source/需求.md`, and structured `allowedPaths`.",
-        "- Only high-confidence frontend implementation demands select the frontend DAG automatically; backend, mixed, negated, and unknown demands keep the template selected by the normal governance profile.",
+        "- standard tasks are classified conservatively by combining the task title, `source/需求.md`, structured `allowedPaths`, strong React/Next/Vue project evidence, and user-visible interaction/state semantics.",
+        "- An implicit frontend demand selects the frontend DAG only when project or frontend-path evidence is paired with product-delivery semantics; framework evidence alone is insufficient. Backend, mixed, negated, documentation/test maintenance, and unknown demands keep the template selected by the normal governance profile.",
         "- `frontend-mock-assess-pi` reads Mock/API/schema evidence and selects `native|browser-intercept|request-adapter|not-needed|blocked` before frontend planning; its deterministic gate rejects blocked or malformed output, while the existing frontend implementer remains the only writer.",
         "- Mock-backed verification never proves real API integration. When the backend was not exercised, closeout must retain the gap and name `<task-id>-real-api-integration-verify`; that follow-up is explicitly created/run after backend readiness, never automatic.",
         "- Automatic frontend classification never replaces a supervised template selected by an explicit profile, workflowPolicy, or the supervised quality gate.",

package/dist/executors/shell-executor.js

@@ -17,7 +17,7 @@ import { formatFrontendReviewContextStdout, runFrontendReviewContextGate, } from
 import { materializeFrontendLintAssessment, materializeFrontendLintBaseline, } from "../workflows/dag/frontend-lint-baseline.js";
 import { formatTraceabilityGateStdout, materializeBackendTestCaseManifest, runBackendTestTraceabilityGate, } from "../workflows/dag/backend-test-case-manifest.js";
 import { materializeBackendTestExecutionContract } from "../workflows/dag/backend-test-execution-contract.js";
-import { materializeBackendTestResultFromRunDir, parseJunitXml } from "../workflows/dag/backend-test-result-contract.js";
+import { materializeBackendTestResultFromRunDir, parsePytestHtmlReport } from "../workflows/dag/backend-test-result-contract.js";
 import { collectBackendTestHumanCaseCatalog, collectBackendTestMappedPytestScripts, hasBlockingBackendMarkdownSafetyFindings, inspectBackendTestEnvironment, requiredBackendMarkdownCaseAcIds, renderBackendTestFacts, renderBackendTestHtml, redactBackendTestOutput, validateBackendMarkdownCases, validateBackendMarkdownTraceability, writeRunReport, } from "../workflows/dag/backend-test-markdown-workflow.js";
 import { buildBackendTestCanonicalResultFromInitialShellSnippet, materializeBackendTestClassification, } from "../workflows/dag/backend-test-classification-contract.js";
 import { backendTestSemanticReviewSchema, materializeBackendTestSemanticReview, } from "../workflows/dag/backend-test-semantic-review-contract.js";
@@ -102,7 +102,12 @@ export async function executeShellCommand(input) {
         const startedAt = Date.now();
         const injectedEnv = input.dagRunMeta
             ? {
-                HARNESS_DAG_RUN_DIR: input.dagRunMeta.runDir,
+                // Git Bash treats backslashes as escape characters inside double
+                // quotes. On Windows, path.join() produces backslash runDirs like
+                // D:\\...\\active\\<runId>, which corrupt "${HARNESS_DAG_RUN_DIR}/..."
+                // expansions. Normalize to POSIX separators so bash sees a stable
+                // path regardless of host platform.
+                HARNESS_DAG_RUN_DIR: input.dagRunMeta.runDir.replaceAll(path.sep, "/"),
                 HARNESS_DAG_RUN_ID: input.dagRunMeta.runId,
             }
             : undefined;
@@ -384,10 +389,18 @@ async function executeBackendTestPipeline(input, meta) {
                 'mkdir -p "${HARNESS_DAG_RUN_DIR}/reports"',
                 'PYTHON_BIN="$(command -v python || command -v python3 || true)"',
                 'if [ -z "${PYTHON_BIN}" ]; then echo "python/python3 is required for backend-test execution" >&2; exit 127; fi',
-                `PYTHONUTF8=1 PYTHONIOENCODING=utf-8 PYTHONDONTWRITEBYTECODE=1 "\${PYTHON_BIN}" -m pytest ${pytestTargets} -v -p no:cacheprovider -o junit_logging=all -o junit_log_passing_tests=true --junitxml="\${HARNESS_DAG_RUN_DIR}/reports/backend-test.junit.xml"`,
+                // Run pytest exactly once and emit the native pytest-html
+                // self-contained report. The HTML/facts renderers read per-case
+                // captured stdout (HTTP_REQUEST/HTTP_RESPONSE) directly from the
+                // pytest-html data-jsonblob island, so a JUnit XML report is no
+                // longer generated.
+                `PYTHONUTF8=1 PYTHONIOENCODING=utf-8 PYTHONDONTWRITEBYTECODE=1 "\${PYTHON_BIN}" -m pytest ${pytestTargets} -v -p no:cacheprovider --html="\${HARNESS_DAG_RUN_DIR}/reports/backend-test.html" --self-contained-html`,
                 "STATUS=$?",
                 'printf "%s" "${STATUS}" > "${HARNESS_DAG_RUN_DIR}/reports/backend-test-pytest-exit.txt"',
-                'if { [ "${STATUS}" -eq 0 ] || [ "${STATUS}" -eq 1 ]; } && [ -s "${HARNESS_DAG_RUN_DIR}/reports/backend-test.junit.xml" ]; then exit 0; fi',
+                // exit 0 (all pass) or 1 (assertion failures) with a valid
+                // pytest-html report are reportable; collection errors / crashes
+                // (exit >= 2) or a missing report surface as a pipeline error below.
+                'if { [ "${STATUS}" -eq 0 ] || [ "${STATUS}" -eq 1 ]; } && [ -s "${HARNESS_DAG_RUN_DIR}/reports/backend-test.html" ]; then exit 0; fi',
                 'exit "${STATUS}"',
             ].join("; ");
             const results = await executePipelineCommands(input, meta, [pytestCommand]);
@@ -396,11 +409,11 @@ async function executeBackendTestPipeline(input, meta) {
                 return { ok: false, stdout: results.map((result) => result.stdout).join("\n"), stderr: failure.stderr, failureCategory: failure.failureCategory, durationMs: Date.now() - started };
             }
             const reportsDir = path.join(meta.runDir, "reports");
-            const junitContent = await readFile(path.join(reportsDir, "backend-test.junit.xml"), "utf8");
+            const pytestHtmlContent = await readFile(path.join(reportsDir, "backend-test.html"), "utf8");
             const pytestExitCode = Number.parseInt((await readFile(path.join(reportsDir, "backend-test-pytest-exit.txt"), "utf8")).trim(), 10);
             if (![0, 1].includes(pytestExitCode))
                 throw new Error(`pytest did not complete with a reportable exit code: ${pytestExitCode}`);
-            const parsed = parseJunitXml(junitContent);
+            const parsed = parsePytestHtmlReport(pytestHtmlContent);
             const cases = await collectBackendTestHumanCaseCatalog(input.cwd);
             const caseValidationSummary = await readRequiredRunReport(reportsDir, "backend-md-case-validation.md");
             const traceabilitySummary = await readRequiredRunReport(reportsDir, "backend-test-traceability.md");
@@ -413,7 +426,7 @@ async function executeBackendTestPipeline(input, meta) {
                 traceabilitySummary,
             });
             const htmlPath = await writeRunReport(meta.runDir, "backend-test.html", htmlContent);
-            const facts = renderBackendTestFacts({ parsed, cases, pytestExitCode, junitRelativePath: "reports/backend-test.junit.xml", htmlRelativePath: "reports/backend-test.html", junitContent, htmlContent, caseValidationSummary, traceabilitySummary });
+            const facts = renderBackendTestFacts({ parsed, cases, pytestExitCode, htmlRelativePath: "reports/backend-test.html", htmlContent, caseValidationSummary, traceabilitySummary });
             const markdownPath = await writeRunReport(meta.runDir, "backend-test.md", facts);
             const factsPath = await writeRunReport(meta.runDir, "backend-test-facts.md", facts);
             const sanitizedOutputs = results.map((result) => redactBackendTestOutput(result.stdout));