fix(web): 进度条自愈——回放层强制单调完成(d1285247 复盘,bump 0.36.2)
task_progress 回放非渲染 bug:模型跳步推进时漏给上一步补 completed, 导致"下面绿勾、上面红圈"。progress.js 加 enforceMonotonicProgress: 某步 completed 则其之前所有步自动 completed,set_plan/update_step 出口 各过一遍,漏发自愈。前端单测 +3(含复刻 d1285247 跳步序列→6/6)。 诊断脚本 scripts/diag_progress_d1285247.py。 Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
This commit is contained in:
parent
fcc158dff6
commit
a89c7386fd
|
|
@ -21,6 +21,9 @@
|
|||
|
||||
## 已完成关键能力
|
||||
|
||||
### 2026-07-03 / 进度条自愈:回放层强制单调完成(d1285247 复盘,bump 0.36.2)
|
||||
用户报 task d1285247(ppt生成3)进度条反常:后面步(质检/导出)打绿勾、前面步(摄取素材/配图)却卡红圈"…",顶部"4/6"。诊断脚本 `scripts/diag_progress_d1285247.py` 拉出 `task_progress` 调用序列定位**非渲染 bug**——`progress.js` 忠实回放了模型发的调用:模型每次推进是"标下一步 completed + 再下一步 in_progress"的跳步,**每次都漏给上一次留在 in_progress 的那步补 completed**(s1、s3 被漏),回放到最后就是 `s1=in_progress,s2=completed,s3=in_progress,s4/s5/s6=completed`。根因是模型用工具收尾不稳,纯提示拦不住(与门体系教训同构)。修在**回放层加确定性单调不变量**:`enforceMonotonicProgress`——checklist 线性推进,只要某步 completed,其之前所有步自动视为 completed;`applyProgressAction` 的 set_plan / update_step 两条出口都过一遍,漏发自愈。前端单测加 3 条(含复刻 d1285247 跳步序列 → 6/6)。已知边界:假设步骤线性顺序(现有所有 skill 成立);若将来出现真·并行/乱序 checklist 会被抹平。
|
||||
|
||||
### 2026-07-03 / ppt 门体系二轮硬化:逃生口收紧 + 导出自动质检 + svg_final 嵌图修复(139a59c5 重跑复盘,bump 0.36.1)
|
||||
0.36.0 上线后同 task 重跑(仍 deepseek-v4-flash):产物整体大幅好转,但仍有 4/25 页错位(P12 色带裁两行标题+正文跑出卡外 / P14·P18 文字骑卡片边框 / P21 手画饼图弧线劈叉)。轨迹显示**两道新门都触发了、都被模型 8 秒内用逃生口按过去**:质检+渲图验收 0 调用,`--allow-iconless` + `--allow-unreviewed` 连按直接导出——门有了,逃生口对弱模型等于"报错时该加的参数"。且 `--allow-iconless` 的"正当理由"是我们自己给的:wrapper docstring 老示例教它 `-s final`,而图标门检查的是 svg_final(data-icon 已展开)→ 误报零图标;`-s final` 还连锁出图片路径连环坑(见 F)。二轮修五处:**A 验收门分层**——"从没渲过/渲后又改/finalize 前渲的"为硬问题,**任何 CLI flag 不豁免**(渲图便宜且机器可验,没理由交付没人能看过的页);`--allow-unreviewed` 只豁免"渲过但没标 pass";运维兜底走 `ZCBOT_PPT_FORCE_EXPORT=1` 环境变量(不进 --help/SKILL)。**B 拔 `-s final` 雷**——图标门永远对 svg_output 源检测(误报根除);wrapper docstring 示例去掉 `-s final` 并注明勿用。**C 导出自动质检门**——svg_to_pptx 导出前内嵌复跑 quality checker 逐页硬错误(坏 XML/禁用特性/图片缺失/几何 error),error 拒绝导出、无豁免参数(fail-open 于 import 失败)——"忘跑/不跑质检"从此无效。**D** 验收门报错计数措辞修正。**E 几何质检加"文字骑卡片边缘"检测**(warning 带坐标:文字与可见矩形交叠面积占比 0.2–0.85 即骑边,P12/P14/P18 三类当场可命中;P21 饼图弧线错误静态无解,只能渲图过目)。**F 修 svg_final 嵌图失效 bug**——finalize 先 copytree 到 `.build/svg_final` 再就地嵌图,`../images/` 从 svg_final 解析必落空 → **所有 deck 的 svg_final 一直嵌不进外链图**(渲图验收 PNG 里图片也是空的);`_resolve_image_path` 加"rebase 回 svg_output 同相对路径"兜底,实测 data:URI 落位。本机全链路回归:未渲→硬拒(带 flag 也拒)/ pending→拒、flag 放 / pass→放行 / 质检 error→拒 / env 强制→放;71 charts 模板几何 0 error。已知边界:P21 类"图形画错但不重叠不越界"仍只有渲图过目能拦——"看没看"无法机器验证,治本要平台层 vision 验收(待做,同 0.35.1 备注)。
|
||||
|
||||
|
|
|
|||
|
|
@ -1,3 +1,3 @@
|
|||
# zcbot 版本号单一事实源:web/app.py 的 FastAPI version、/healthz 返回、前端展示都引这里。
|
||||
# 改版本只动这一行。
|
||||
__version__ = "0.36.1"
|
||||
__version__ = "0.36.2"
|
||||
|
|
|
|||
|
|
@ -0,0 +1,56 @@
|
|||
"""Dump the task_progress tool-call sequence for a task (by id prefix). ASCII-only."""
|
||||
import json
|
||||
import os
|
||||
import sys
|
||||
from pathlib import Path
|
||||
|
||||
env = Path(__file__).resolve().parent.parent / ".env"
|
||||
for line in env.read_text(encoding="utf-8").splitlines():
|
||||
if line.strip().startswith("ZCBOT_DB_URL="):
|
||||
os.environ["ZCBOT_DB_URL"] = line.split("=", 1)[1].strip()
|
||||
from sqlalchemy import create_engine, text # noqa: E402
|
||||
|
||||
engine = create_engine(os.environ["ZCBOT_DB_URL"])
|
||||
prefix = sys.argv[1] if len(sys.argv) > 1 else "d1285247"
|
||||
|
||||
with engine.connect() as conn:
|
||||
row = conn.execute(
|
||||
text("select task_id,name,status,run_status from tasks where task_id::text like :p"),
|
||||
{"p": prefix + "%"},
|
||||
).fetchone()
|
||||
if not row:
|
||||
print("[NO TASK]", prefix)
|
||||
sys.exit(1)
|
||||
tid = row[0]
|
||||
print(f"[TASK] {tid} name={row[1]!r} status={row[2]} run={row[3]}")
|
||||
|
||||
msgs = conn.execute(
|
||||
text("select idx,payload from messages where task_id=:t order by idx"),
|
||||
{"t": tid},
|
||||
).fetchall()
|
||||
print(f"[MESSAGES] {len(msgs)}")
|
||||
|
||||
n = 0
|
||||
for idx, p in msgs:
|
||||
for tc in p.get("tool_calls") or []:
|
||||
fn = tc.get("function") or {}
|
||||
if fn.get("name") != "task_progress":
|
||||
continue
|
||||
n += 1
|
||||
try:
|
||||
args = json.loads(fn.get("arguments") or "{}")
|
||||
except Exception as e:
|
||||
print(f" [{idx}] PARSE-ERR: {e} raw={fn.get('arguments')!r}")
|
||||
continue
|
||||
act = args.get("action")
|
||||
if act == "set_plan":
|
||||
steps = args.get("steps") or []
|
||||
print(f" [{idx}] set_plan ({len(steps)} steps):")
|
||||
for st in steps:
|
||||
print(f" {st.get('id')!r:8} {st.get('status'):11} {st.get('title')!r}")
|
||||
elif act == "update_step":
|
||||
st = args.get("step") or {}
|
||||
print(f" [{idx}] update_step id={st.get('id')!r} status={st.get('status')!r} title={st.get('title')!r}")
|
||||
else:
|
||||
print(f" [{idx}] {act} {json.dumps(args, ensure_ascii=False)}")
|
||||
print(f"[task_progress calls] {n}")
|
||||
|
|
@ -3,6 +3,7 @@ import test from "node:test";
|
|||
|
||||
import {
|
||||
applyProgressAction,
|
||||
enforceMonotonicProgress,
|
||||
progressActionsFromToolCalls,
|
||||
} from "../web/static/js/progress.js";
|
||||
|
||||
|
|
@ -49,3 +50,70 @@ test("tool calls can apply progress updates on top of previous task progress", (
|
|||
{ id: "s2", title: "实现功能", status: "pending" },
|
||||
]);
|
||||
});
|
||||
|
||||
test("a completed step force-completes earlier dangling steps (monotonic heal)", () => {
|
||||
const steps = [
|
||||
{ id: "s1", title: "摄取素材", status: "in_progress" },
|
||||
{ id: "s2", title: "策略", status: "completed" },
|
||||
{ id: "s3", title: "配图", status: "pending" },
|
||||
];
|
||||
assert.deepEqual(enforceMonotonicProgress(steps), [
|
||||
{ id: "s1", title: "摄取素材", status: "completed" },
|
||||
{ id: "s2", title: "策略", status: "completed" },
|
||||
{ id: "s3", title: "配图", status: "pending" },
|
||||
]);
|
||||
});
|
||||
|
||||
test("update_step marking a later step completed heals the earlier in_progress step", () => {
|
||||
const initial = applyProgressAction([], {
|
||||
action: "set_plan",
|
||||
steps: [
|
||||
{ id: "s1", title: "摄取素材", status: "in_progress" },
|
||||
{ id: "s2", title: "策略", status: "pending" },
|
||||
{ id: "s3", title: "配图", status: "pending" },
|
||||
],
|
||||
});
|
||||
|
||||
const updated = applyProgressAction(initial, {
|
||||
action: "update_step",
|
||||
step: { id: "s2", status: "completed" },
|
||||
});
|
||||
|
||||
assert.deepEqual(updated, [
|
||||
{ id: "s1", title: "摄取素材", status: "completed" },
|
||||
{ id: "s2", title: "策略", status: "completed" },
|
||||
{ id: "s3", title: "配图", status: "pending" },
|
||||
]);
|
||||
});
|
||||
|
||||
// Replay the exact d1285247 task_progress sequence that produced the reported
|
||||
// "green check below, red dot above" bug; the heal must yield a clean 6/6.
|
||||
test("replays the d1285247 skip-ahead sequence to a fully completed plan", () => {
|
||||
const call = (args) => ({
|
||||
function: { name: "task_progress", arguments: JSON.stringify(args) },
|
||||
});
|
||||
const seq = [
|
||||
call({
|
||||
action: "set_plan",
|
||||
steps: [
|
||||
{ id: "s1", title: "摄取素材", status: "in_progress" },
|
||||
{ id: "s2", title: "策略", status: "pending" },
|
||||
{ id: "s3", title: "配图", status: "pending" },
|
||||
{ id: "s4", title: "执行", status: "pending" },
|
||||
{ id: "s5", title: "质检", status: "pending" },
|
||||
{ id: "s6", title: "导出", status: "pending" },
|
||||
],
|
||||
}),
|
||||
call({ action: "update_step", step: { id: "s2", status: "completed" } }),
|
||||
call({ action: "update_step", step: { id: "s3", status: "in_progress" } }),
|
||||
call({ action: "update_step", step: { id: "s4", status: "completed" } }),
|
||||
call({ action: "update_step", step: { id: "s5", status: "in_progress" } }),
|
||||
call({ action: "update_step", step: { id: "s5", status: "completed" } }),
|
||||
call({ action: "update_step", step: { id: "s6", status: "completed" } }),
|
||||
];
|
||||
|
||||
const { steps } = progressActionsFromToolCalls(seq, []);
|
||||
assert.deepEqual(steps.map(s => s.status), [
|
||||
"completed", "completed", "completed", "completed", "completed", "completed",
|
||||
]);
|
||||
});
|
||||
|
|
|
|||
|
|
@ -18,13 +18,30 @@ export function normalizeProgressStep(step) {
|
|||
return { id, title, status };
|
||||
}
|
||||
|
||||
// The checklist is a linear progress bar: work advances top-to-bottom. Models
|
||||
// don't always send a `completed` update before moving the in_progress marker
|
||||
// on (observed: later steps marked done while earlier ones dangle at
|
||||
// in_progress), which renders as "green check below, red dot above". Enforce
|
||||
// monotonic completion — any step before the last completed one is completed
|
||||
// too — so a missed update self-heals instead of stranding earlier steps.
|
||||
export function enforceMonotonicProgress(steps) {
|
||||
if (!Array.isArray(steps)) return [];
|
||||
let lastCompleted = -1;
|
||||
for (let i = 0; i < steps.length; i++) {
|
||||
if (steps[i] && steps[i].status === "completed") lastCompleted = i;
|
||||
}
|
||||
if (lastCompleted <= 0) return steps.map(s => ({ ...s }));
|
||||
return steps.map((s, i) => (i < lastCompleted ? { ...s, status: "completed" } : { ...s }));
|
||||
}
|
||||
|
||||
export function applyProgressAction(progress, args) {
|
||||
const current = cloneProgressSteps(progress);
|
||||
if (!args || typeof args !== "object") return current;
|
||||
const action = args.action || "";
|
||||
if (action === "clear") return [];
|
||||
if (action === "set_plan") {
|
||||
return Array.isArray(args.steps) ? args.steps.map(normalizeProgressStep).filter(Boolean) : [];
|
||||
const planned = Array.isArray(args.steps) ? args.steps.map(normalizeProgressStep).filter(Boolean) : [];
|
||||
return enforceMonotonicProgress(planned);
|
||||
}
|
||||
if (action === "update_step") {
|
||||
const raw = args.step;
|
||||
|
|
@ -48,7 +65,7 @@ export function applyProgressAction(progress, args) {
|
|||
status: normalizeProgressStatus(raw.status),
|
||||
});
|
||||
}
|
||||
return next;
|
||||
return enforceMonotonicProgress(next);
|
||||
}
|
||||
return current;
|
||||
}
|
||||
|
|
|
|||
Loading…
Reference in New Issue