@tea-agent/loop-agent 0.29.0 → 0.29.2

package/docs/templates/README.md

@@ -48,6 +48,7 @@
   Pi writer writeSet 不包含 final manifest（文件名 manifest.json）。依赖 checklist 的 exclusive shell materializer 是 final manifest 的唯一写入者，以 temp+rename 原子生成 final，后续 map 只消费该 materializer 输出。
 - `frontend-test-dag.retrieve-context.prompt.md`、`frontend-test-dag.generate-cases.prompt.md`、`frontend-test-dag.review-cases.prompt.md`、`frontend-test-dag.review-execution.prompt.md`、`frontend-test-dag.retrospect.prompt.md` — 上下文、用例、审查、执行审查和复盘提示。
 - `frontend-test-case-checklist.md` — 前端测试用例检查清单。
+- `frontend-test-standard-scenarios.v1.json` — frontend-test 标准场景规范（generate 覆盖 must 场景；运行时可物化到目标项目 testcase/frontend/rag 下的 standard-scenarios.v1.json）。
 
 ## 知识同步与配置
 

package/docs/templates/agent-dag.schema.json

@@ -56,7 +56,12 @@
           "type": "array",
           "items": { "type": "string", "minLength": 1 }
         },
-        "writePolicy": { "$ref": "#/$defs/writePolicy" }
+        "writePolicy": { "$ref": "#/$defs/writePolicy" },
+        "writeGuardPolicy": {
+          "type": "string",
+          "enum": ["full", "tools-only"],
+          "description": "full (default): git baseline + post-diff write-guard. tools-only: keep tool sandbox only (frontend-test)."
+        }
       }
     },
     "skillsByRole": {

package/docs/templates/frontend-test-case-checklist.md

@@ -7,7 +7,7 @@ LLM review (when `frontendTest.reviewMode=blocking`) must not invent blocking ru
 
 | ruleId | Rule |
 |---|---|
-| `open-prefix` | Each case body includes `playwright-cli open --browser=chrome --headed <absolute-http(s)-url>` |
+| `open-prefix` | Each case body includes `playwright-cli open --browser=chrome --headless <absolute-http(s)-url>` |
 | `production-url` | Open URL must not look like a production host |
 | `ac-mapping` | Manifest entry has non-empty `acIds` (**acceptance** ids, not case ids) |
 | `ac-id-shape` | Each `acIds[]` entry matches `AC-*` / `AC-FE-*` |

package/docs/templates/frontend-test-dag.generate-cases.prompt.md

@@ -18,7 +18,7 @@ Each case is independently executable and includes AC mapping (`acIds`), precond
 
 Every case must copy the **controller-frozen** absolute base URL supplied by the DAG. `testcase/frontend/rag/context.md` may reference that value, but model-authored context/case prose cannot establish or override the origin. Do not leave a base-url placeholder.
 
-Every case must start with `playwright-cli open --browser=chrome --headed` followed by the concrete controller-frozen URL supplied by the DAG. Do not copy an angle-bracket URL placeholder into a generated executable case line.
+Every case must start with `playwright-cli open --browser=chrome --headless` followed by the concrete controller-frozen URL supplied by the DAG. Do not copy an angle-bracket URL placeholder into a generated executable case line.
 
 ### Dynamic element refs (hard)
 
@@ -53,3 +53,11 @@ Each case must require the executor to persist, even when blocked:
 - `testcase/frontend/evidence/<case-id>/case-result.json`
 
 `case-result.json` must be valid JSON containing the matching `caseId`, `status` (`passed`, `failed`, or `blocked`), and an `evidencePaths` array. A blocked result must include a non-empty `blockedReason`, such as `isolated-test-environment-unavailable` or `token-budget-exhausted`, and must never claim or imply a pass. `execution.md` records attempted or blocked steps, base-URL safety decision, fixture/reset and request-observation availability, timestamps, and the evidence-file list. Available evidence files are required only when actually produced and must stay under the same case evidence directory.
+
+
+## Standard scenarios (controller)
+
+- Read `testcase/frontend/rag/standard-scenarios.v1.json` when present.
+- Cover each `priority=must` scenario with at least one case or record an explicit GAP in coverage-map.
+- Each case Markdown SHOULD include `## 测试点` and `## 测试步骤` sections.
+- Prefer linking cases via `standardScenarioIds` in manifest.draft notes when possible.