Commit Graph

2106 Commits

Author SHA1 Message Date
1445043649 54cb94c30f feat(auth): JWT TokenPayload accepts wid + role
TokenPayload gains optional wid (workspace_id) and role claims;
create_access_token accepts them as keyword-only args and only
encodes them when provided. Existing tokens and existing callers
keep working unchanged — the contract that protected requests
must carry wid is enforced by middleware (PR4 T4.6/T4.7), not by
the JWT type system.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-12 22:17:55 +08:00
1445043649 54762f491c feat(persistence): UserRow.default_workspace_id
Adds a nullable FK column to mirror alembic 0001. On fresh deployments
metadata.create_all() will create users with this column; existing
deployments rely on the alembic migration to add it.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-12 22:15:05 +08:00
1445043649 917d8fbeaf test(persistence): alembic 0001 round-trip on SQLite + Postgres
Verifies alembic upgrade 0001 adds users.default_workspace_id (with FK
to workspaces) and that downgrade cleanly removes it. Runs on both
dialects because the migration uses op.batch_alter_table for SQLite
ALTER compatibility.

Tests are intentionally sync — alembic's command layer is sync and
env.py calls asyncio.run(); running under pytest-anyio would deadlock
on the inner event loop.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-12 22:11:11 +08:00
1445043649 8efbb2f9e5 feat(persistence): alembic 0001 — users.default_workspace_id
First Alembic revision for the DeerFlow application schema. Adds a
nullable users.default_workspace_id column with an ON DELETE SET NULL
FK to workspaces(id), so newly-registered users can be routed back to
their default workspace on next login without consulting memberships.

Uses op.batch_alter_table for SQLite ALTER compatibility (sqlite is
still a supported dev fallback).

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-12 22:08:44 +08:00
1445043649 d98498b705 docs(impl): T4.1 — confirm alembic baseline is not needed
alembic heads/history both empty, versions/ dir empty: 0001 can be the
first revision with down_revision=None. alembic_version table will be
auto-created on first upgrade head. doctor.py needs no new check.

Also records T4 preparation grep findings on _ensure_admin_user current
behavior (open issue #3 resolved).

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-12 22:07:34 +08:00
1445043649 c70c6594de docs(impl): STATUS — record LOCK sign-off and origin push
Two unblocking events on 2026-05-12: team signed off all 7 irreversible
schema decisions, and docs branch was finally pushed to origin (38 commits,
via SSH-over-443 to bypass local proxy). PR4 is now unblocked.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-12 22:02:33 +08:00
1445043649 a592319e3c docs(impl): STATUS update after PR3 merge
PR3  merged 进 docs branch(7 commits, T3.1-T3.10)+ RDS live 验证通过
(11 张表含 workspaces/workspace_memberships,partial unique on owner 索引
建出来)。测试基线 3134 passed + 25 skipped + 0 failed。

下一个 PR4(auth 改造)plan 推荐 Inline 模式(跨 jwt.py/auth_middleware.py/
routers/auth.py 多文件耦合紧)。

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-12 21:34:05 +08:00
1445043649 dda8264057 docs(impl): PR3 implementation note + acceptance checklist
记录 Stage 0 PR3 的实施落点、LOCK 决策、跟进项与 6 commit 列表。

验收:3134 passed + 25 skipped + 0 failed(PR2 末 3087 + 47 新 ws_context/
ws_repo/membership_repo 测试 + 2 PG-only skipped)。RDS live 验证通过:
workspaces / workspace_memberships 表 + partial unique on owner 都建出来。

T3.6 (membership ORM) 提前到 T3.4-T3.5 (workspace repo) 之前,因为
WorkspaceRepository.get/list_by_user 会 JOIN memberships。

Stage 0 PR3 T3.10. PR3 完成。

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-12 21:18:19 +08:00
1445043649 3313047ff1 test(workspace): partial unique on owner — Postgres twin test
新建 test_workspace_partial_unique_postgres.py,2 个 @pytest.mark.postgres
test,用 PR1 的 testcontainers postgres_url fixture 验:
  - test_partial_unique_on_owner_enforced_on_postgres:第 2 个 owner
    IntegrityError(SQLite twin 在 test_workspace_membership_repo 已覆盖)
  - test_multiple_admins_allowed_on_postgres:多个 admin/member 不触发约束

为什么 PG 单独写:sqlite_where vs postgresql_where 是两份 DDL;SQLAlchemy
能 emit 不代表 PG 真的执行。本测试 pin 两边语义一致。

本地无 docker daemon → SKIPPED;CI workflow backend-postgres-tests 会实跑。

Stage 0 PR3 T3.8。

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-12 21:15:25 +08:00
1445043649 36ffe2713f feat(persistence): WorkspaceMembershipRepository + 8 unit tests
新建 backend/packages/harness/deerflow/persistence/workspace_membership/sql.py,方法:
  - add(workspace_id, user_id, role, invited_by=None) 校验 role ∈
    {owner, admin, member}(Stage 0 仅写 owner,schema 已支持其余)
  - remove(workspace_id, user_id) → bool(rowcount > 0)
  - list_by_user(user_id) 按 joined_at desc
  - list_by_workspace(workspace_id) 按 joined_at asc
  - get_role(workspace_id, user_id) → str | None
  - change_role(workspace_id, user_id, new_role) → bool

MembershipValidationError 自定义异常(role 非法)。

8 test 覆盖:
  - add → get_role round-trip
  - remove 命中/未命中返回 True/False
  - 同 workspace 第 2 个 owner 触发 IntegrityError(partial unique)
  - 同 workspace 多个 admin 不触发(Stage 2 forward compat)
  - CASCADE: 删 user 自动清理 memberships
  - list_by_user 按 joined_at desc 排序(多 workspace)
  - role 校验拒绝 'viewer'
  - change_role 命中改值 + 未命中返 False

注:owner 转让需要事务内两行原子 swap(先把现 owner 改 admin,再把新 owner
插 owner),放 PR4+ auth router 内做带权限检查的版本;本仓储不暴露
transfer_ownership 方法以保持单一职责。

Stage 0 PR3 T3.7。

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-12 21:13:52 +08:00
1445043649 a40df03521 feat(persistence): WorkspaceRepository + 23 unit tests
新建 backend/packages/harness/deerflow/persistence/workspace/sql.py,方法:
  - create(name, slug, owner_id, *, workspace_id=None, status='active')
    UUID v4 自动生成;强校验 slug 格式(regex ^[a-z0-9](-?[a-z0-9])*\$ + 3-32
    长度)+ slug 黑名单(25 个保留字)+ status 枚举
  - get(workspace_id, *, user_id=AUTO) JOIN workspace_memberships 做成员校验;
    user_id=None 显式 bypass(迁移/admin)
  - get_by_slug(slug) 不带成员校验(path-based routing 用:先 slug→workspace_id
    再到 route handler 里查成员)
  - list_by_user(*, user_id=AUTO) 列 user 所属所有 workspace
  - update_status / delete platform-admin 操作,不带成员校验

WorkspaceValidationError 自定义异常(slug 格式 / 黑名单 / status)。

23 test 覆盖:
  - CRUD smoke + get_by_slug missing
  - 重复 slug → IntegrityError
  - 8 个 invalid slug pattern(短/长/大写/空格/破折号位置/连续破折号/下划线)
  - 6 个 blacklisted slug
  - status 状态机 + 非法值拒绝
  - delete CASCADE 到 memberships(SQLite FK PRAGMA 已开启)
  - get/list 成员过滤(user-A 看不见 user-B 的 workspace)
  - list user_id=None 显式 bypass

全部在 SQLite ephemeral DB 上跑(< 1s)。partial-unique 双驱动验证留给 T3.8。

Stage 0 PR3 T3.4 + T3.5。

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-12 21:11:28 +08:00
1445043649 d2d2d29c34 feat(persistence): add WorkspaceMembershipRow ORM model
新建 backend/packages/harness/deerflow/persistence/workspace_membership/
{__init__.py, model.py},注册到 Base.metadata。

WorkspaceMembershipRow schema(来自 workspace-schema-design §2.2 锁定版):
  - workspace_id String(36) FK workspaces.id ON DELETE CASCADE — PK part
  - user_id String(36) FK users.id ON DELETE CASCADE — PK part
  - role String(16) NOT NULL(Stage 0 仅写 'owner',schema 允许 'admin'/'member'
    为 Stage 2 RBAC 准备)
  - invited_by String(36) FK users.id ON DELETE SET NULL nullable(Stage 2
    invitation 流程用)
  - joined_at DateTime(tz=True)

索引:
  - idx_workspace_memberships_user (user_id, workspace_id) — 倒查索引让
    /auth/me 列 user 所属 workspaces 走索引
  - idx_one_owner_per_workspace partial UNIQUE on workspace_id WHERE role='owner'
    —— 一 workspace 严格 1 个 owner;sqlite_where + postgresql_where 双驱动
    并存(实测两边都识别)

Stage 0 PR3 T3.6(提前到 T3.4-T3.5 之前;WorkspaceRepository.list_by_user
等会 JOIN 这张表)。

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-12 21:09:19 +08:00
1445043649 8323bf68d2 feat(persistence): add WorkspaceRow ORM model
新建 backend/packages/harness/deerflow/persistence/workspace/{__init__.py, model.py}
+ 在 persistence.models.__init__ 注册 WorkspaceRow 让 Base.metadata.create_all
能自动建表。

WorkspaceRow schema(来自 workspace-schema-design §2.1 锁定版):
  - id String(36) PK(UUID v4 字符串,与 users.id 对齐)
  - name String(64) NOT NULL(显示名)
  - slug String(32) NOT NULL UNIQUE(URL 标识,正则 ^[a-z0-9](-?[a-z0-9])*$)
  - status String(16) default 'active'(active/suspended/deleted)
  - owner_id String(36) FK users.id ON DELETE RESTRICT(删 owner 时阻拦,
    必须先转让所有权)
  - created_at / updated_at DateTime(tz=True)

每列 + 表都带中文 comment(沿用 commit 9ff79055 的 convention)。

Stage 0 PR3 T3.3。

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-12 21:07:30 +08:00
1445043649 f63089aea8 feat(runtime): add workspace_context module + tests
新建 backend/packages/harness/deerflow/runtime/workspace_context.py,仿
user_context.py 的 API 形态:

  - CurrentWorkspace Protocol(要求 .id: str + .role: str)
  - _current_workspace ContextVar + set/reset/get/require
  - DEFAULT_WORKSPACE_ID = "default" + get_effective_workspace_id(fallback
    友好,不抛错;用于文件系统路径)
  - AUTO 哨兵 + resolve_workspace_id 三态(AUTO/str/None)

与 user_context 的区别:CurrentWorkspace 额外要求 .role 字段,让 Protocol 同时
约束"workspace 是哪个"和"caller 在该 workspace 内的角色"(Stage 0 只见 'owner',
Stage 2 RBAC 打开 admin/member)。

backend/tests/test_workspace_context.py 16 个 test,覆盖:
  - 4 个 set/reset/require 行为
  - 3 个 Protocol structural check(接受 .id+.role / 拒少 .role / 拒少 .id)
  - 4 个 get_effective_workspace_id(含 UUID → str 强转)
  - 5 个 resolve_workspace_id 三态(AUTO/AUTO 无 ctx 抛错/explicit str/explicit
    None/AUTO 强转 str)

Stage 0 PR3 T3.1 + T3.2。

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-12 21:04:49 +08:00
1445043649 cbbb83a706 docs(impl): STATUS update after PR2 live verification on RDS
PR1 + PR2 已 live 验证:远程 Aliyun RDS (PostgreSQL 17.9) 上 9 张表
(DeerFlow 5 + LangGraph 4) 全部 create_all 成功。

记 3 个 PR2 follow-up commits(不在原 plan task list 但 live 必需):
  - testcontainers image 16→17 对齐 RDS 大版本
  - serve.sh auto-add --extra postgres 防 uv sync 卸 asyncpg
  - async_provider 剥 +asyncpg dialect 前缀让 LangGraph saver 能解析 URL

更新"一句话状态" + 标 PG live 验证(区分 testcontainers 路径仍未实跑)。

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-11 18:01:22 +08:00
1445043649 39f8e117e8 fix(checkpointer): strip SQLAlchemy dialect prefix before passing to LangGraph
DeerFlow 用同一个 \`postgres_url\` 喂两条路径:
  - SQLAlchemy 引擎(需 \`postgresql+asyncpg://\`)
  - LangGraph AsyncPostgresSaver(psycopg 直连,需 libpq 风格 \`postgresql://\`)

PR2 把 .env / config.example.yaml 默认 URL 改成 \`+asyncpg\` 形态后,LangGraph
saver from_conn_string 会被 psycopg parser 直接抛 ProgrammingError
'missing "=" after "postgresql+asyncpg://..."'。

修:在 async_provider.py 传给 AsyncPostgresSaver 之前用正则剥掉
\`postgresql+\\w+://\` → \`postgresql://\`。一行修复,让同一个 URL 同时满足
SQLAlchemy 和 LangGraph 两条路径。

实测:make dev-daemon 起 gateway 成功,远程 RDS 上 9 张表(DeerFlow 5 +
LangGraph 4)全部 create_all 出来。

Stage 0 PR2 follow-up(不在原 plan task list 内,是 live make dev 触发的)。

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-11 17:54:57 +08:00
1445043649 bdef6c6b9f fix(serve): auto-install postgres extras when config.yaml selects postgres
scripts/serve.sh 在 \`uv sync --quiet\` 前 grep config.yaml 探测
\`database.backend: postgres\`;若是则追加 \`--extra postgres\`,否则保持原状。

修一个 PR1+PR2 没覆盖到的隐患:serve.sh 每次 dev 启动都跑 uv sync,会把
之前手动 \`uv sync --extra postgres\` 装的 asyncpg 卸掉,导致 gateway
启动时 ImportError。

Stage 0 PR2 follow-up(不在原 plan task list 内,是 live 验证时发现)。

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-11 17:35:23 +08:00
1445043649 7f17cb8fee chore(test): bump testcontainers Postgres to 17-alpine
\`make doctor\` 实测远程 Aliyun RDS = PostgreSQL 17.9。把
backend/tests/fixtures/postgres.py 的 testcontainers 镜像从
postgres:16-alpine 调到 postgres:17-alpine,与生产对齐。

同步把 STATUS.md "RDS 大版本对齐" 项标 done。

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-11 17:29:16 +08:00
1445043649 ad22242ecc docs(impl): add Stage 0 progress dashboard (STATUS.md)
新增 docs/multi-tenant-redesign/03-impl/STATUS.md 作为 Stage 0 唯一的"现在
到哪了"权威来源:

  - 8 PR 状态表(PR0-PR2  merged,PR3-PR8 🟡 pending)
  - 用户必须跟进的事(live PG 实跑 / RDS 大版本对齐 / push origin / LOCK
    review / 远程 RDS 密码轮换)
  - 跳过/推迟的子任务(T1.10 docker-pending、T2.7 acked、T2.8 optional
    skipped、T2.9 backend/CLAUDE.md 待补)
  - 即将遇到的开放问题(PR4 alembic baseline / _ensure_admin_user 现状 /
    PG 大版本)
  - PR3 模式选择 trade-off(Inline vs Subagent-Driven)
  - 维护规则:完成 PR 后必更新;新 session 首先读本文件

填补 plan/impl notes 之外的空白:plan 是静态的,impl notes 是 per-PR
快照;STATUS 是跨 PR 的执行状态视图。

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-11 16:44:09 +08:00
1445043649 1112a1971b docs(impl): PR2 implementation note + acceptance checklist
记录 Stage 0 PR2 的实施落点、关键 LOCK 决策、跟进项与 7 commit 列表。
验收:3087 passed + 23 skipped + 0 failed(vs PR1 基线 3085 + 2 新
default-backend 测试)。

Stage 0 PR2 T2.10. PR2 完成。

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-11 09:23:16 +08:00
1445043649 c53295dfee docs(readme): add Database backend section in Quick Start
Quick Start 加 Step 3 "Database backend (Stage 0+ defaults to Postgres)":
  - 说明 config.example.yaml 默认 postgres + DATABASE_URL 写 .env
  - 给本地 dev 起 docker compose postgres 一行命令
  - 提示 make doctor / make dev preflight 行为
  - <details> 折叠 SQLite fallback 说明(offline dev 用)

Stage 0 PR2 T2.9(README 部分;backend/CLAUDE.md follow-up)。

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-11 09:20:26 +08:00
1445043649 745a33e05d chore: T2.7 acknowledgement (covered by T1.8)
Plan T2.7 = "setup_wizard.py 推荐 postgres,引导填 DATABASE_URL"。
T1.8 commit eae01901 已经实现:

  - 加 "Use Postgres? (y = postgres, n = sqlite)" 问答,默认 y
  - 选 y 时引导填 DATABASE_URL,可留空稍后填 .env
  - writer.build_minimal_config(database_backend='postgres') 输出
    database.backend=postgres + postgres_url=\$DATABASE_URL

无新代码改动;本 commit 仅为 task tracking 完整性。

Stage 0 PR2 T2.7(acknowledged-only)。

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-11 09:17:51 +08:00
1445043649 83b680b2ea feat(check): postgres preflight in scripts/check.py
新增 check_postgres_preflight() —— make dev 链 (check.py → serve.sh) 的最后一道
preflight:
  - config.yaml 不存在 → silent skip(让 setup_wizard 引导)
  - backend != postgres → silent skip
  - DATABASE_URL 未设 → FAIL with hint
  - postgres 设了但 host:port 3s socket 不通 → FAIL with 启 docker 提示
  - 通则 OK + 显示 host:port

不在 serve.sh 里加:Makefile 已经把 check.py 串在 serve.sh 之前,FAIL 会
自然阻断启动;避免 bash + python 两处实现 PG 探测。

doctor.py 的 check_database 是事后诊断(make doctor);本 check 是事前
preflight(make dev/start)—— 互补。

Stage 0 PR2 T2.6.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-11 09:17:19 +08:00
1445043649 3e62a0f6ef feat(docker): gateway depends_on postgres healthcheck (dev only)
dev compose gateway 段加 depends_on: { postgres: { condition: service_healthy } },
确保 postgres 通过 pg_isready 后 gateway 才启 uvicorn。

prod compose 不动:production 用远程 RDS(不在 compose 内),无 postgres
service 可 depend;docker-compose.yaml 不加此 depends_on。

YAML 通过 docker compose config 校验。

Stage 0 PR2 T2.5.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-11 09:13:44 +08:00
1445043649 d312bdf968 test(config): pin default postgres backend + sqlite regression
新建 backend/tests/test_default_database_backend.py 两个测试:

  - test_explicit_sqlite_backend_still_works (T2.3)
    显式 database.backend=sqlite 仍生效;防 PR2 改默认后 SQLite 用户
    悄无声息 regression
  - test_config_example_default_backend_is_postgres (T2.4)
    直接读 on-disk config.example.yaml,断言 database.backend == 'postgres'
    且 postgres_url == '\$DATABASE_URL'(不允许硬编码凭据)

Stage 0 PR2 T2.3 + T2.4.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-11 09:12:28 +08:00
1445043649 7d3d3560ad feat(env): activate DATABASE_URL with local-dev default
.env.example DATABASE_URL 行从注释改为活值(指向 docker compose 起的本地
postgres)。同时加 RDS 示例注释作为参考。psycopg2 风格 URL 升级到
postgresql+asyncpg:// 与 SQLAlchemy 异步 dialect 对齐。

Stage 0 PR2 T2.2.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-11 04:24:32 +08:00
1445043649 404135a16a feat(config): default database backend to postgres
config.example.yaml database 段:postgres 成为活跃默认(postgres_url=\$DATABASE_URL),
SQLite 改为注释掉的 fallback 块(offline dev 用)。bump config_version 9→10。

为什么默认改 postgres:Stage 0+ 已要求 ALTER 4 张表加 workspace_id;
SQLite 加列后再迁 PG 是返工。Stage 0 没有生产数据,迁移阻力最小。

Stage 0 PR2 T2.1.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-10 23:03:52 +08:00
1445043649 85a14f4c05 docs(impl): PR1 implementation note + acceptance checklist
记录 Stage 0 PR1 的实施落点、关键 LOCK 决策、fixture 用法示例、跟进项与
10 commit 列表。验收:3085 passed + 23 skipped + 0 failed(无 regression),
PG smoke 在 docker daemon 起来后实跑(CI workflow 已配)。

Stage 0 PR1 T1.10. PR1 完成。

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-10 22:58:21 +08:00
1445043649 a33b46b4af ci(postgres): add Postgres workflow running @pytest.mark.postgres tests
新增 .github/workflows/backend-postgres-tests.yml,独立于现有
backend-unit-tests workflow:
  - GitHub runner (ubuntu-latest) 自带 docker daemon
  - testcontainers 在 runner 上起 postgres:16-alpine 容器
  - uv sync --extra postgres-test 拉 testcontainers + asyncpg + psycopg
  - pytest -m postgres 仅跑 PG 标签的 4 个 smoke test (+ 后续 PR 增量)

为什么独立 workflow 而不是合到 backend-unit-tests:
  - 现有 fast 测试不受 PG container 启动延迟影响
  - 可独立 fail-soft(早期 Stage 0 rollout 期间需要时加 continue-on-error)
  - fork 用户不强制承担 docker infra 成本

Stage 0 PR1 T1.9.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-10 22:55:53 +08:00
1445043649 eae0190184 feat(wizard): add Postgres backend question to setup wizard
setup_wizard.py 在 Step 3 (Execution) 之后新增一个 Database 问答:
"Use Postgres? (y = postgres, n = sqlite)",默认 y(Stage 0+ 推荐 PG)。
选 y 时引导填 DATABASE_URL(可留空稍后写 .env);DATABASE_URL 进 .env,
config.yaml 写入 database.backend=postgres + postgres_url=\$DATABASE_URL。

writer.py build_minimal_config 加 database_backend 参数,postgres 时
覆盖 base_config 的 database 段;默认 sqlite 时沿用 base_config 行为
(继承 config.example.yaml 的 sqlite_dir 等)。

minimal pattern:不新建 wizard/steps/database.py,inline 在 main 里加 1
问答 + writer 加 1 参数。后续如果需要更复杂数据库选项再升为完整 step 模块。

Stage 0 PR1 T1.8.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-10 22:55:19 +08:00
1445043649 e6f5ba53bd feat(doctor): add Database section probing configured backend
scripts/doctor.py 新增 check_database():

  - sqlite → OK + 显示 sqlite_dir
  - memory → WARN(数据非持久化)
  - postgres → 解析 database.postgres_url 中的 \$DATABASE_URL,asyncpg
    实际连接 + SELECT version() → OK with server version;连接失败时
    FAIL 给出可执行 fix(启 docker compose postgres / 检查 DATABASE_URL)
  - 未知 backend → WARN

Database section 插在 LLM Provider 与 Sandbox 之间。SQLite 部署不会触发
PG 探测(不破坏现有 dev 体验)。

Stage 0 PR1 T1.7.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-10 22:52:26 +08:00
1445043649 8f480cd76f test(persistence): postgres smoke tests for init_engine + ThreadMetaRepo
新建 backend/tests/test_postgres_smoke.py,4 个 test 用 postgres_url fixture:

  - test_postgres_url_creates_isolated_database (T1.4)
    asyncpg 直连验证 fixture 落到 test_<hex> DB
  - test_postgres_url_isolates_between_tests (T1.4)
    手动起第 2 个 DB 比对,证明每 test 隔离
  - test_init_engine_postgres_creates_tables (T1.5)
    init_engine('postgres', postgres_url) 跑完后 information_schema
    出现 5 张现有表(users/threads_meta/runs/feedback/run_events)
  - test_thread_meta_repo_postgres_round_trip (T1.6)
    ThreadMetaRepository.create + get 完整 round-trip,验证 PG 上仓储行为
    与 SQLite 对齐

全部 @pytest.mark.postgres 门控;Docker 不可用时由 fixture 跳过。本次本地
docker daemon 未起,4 测试 SKIP;CI workflow(T1.9)将带 docker service
触发实跑。

Regression:跑全套 `pytest tests/` 3085 passed + 23 skipped + 0 failed
(基线 3086 passed + 18 skipped;diff = 4 新 postgres skip + 1 env 相关
live test 偶发 skip)。

Stage 0 PR1 T1.4 + T1.5 + T1.6.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-10 22:50:10 +08:00
1445043649 31361e2dcf test(fixtures): add testcontainers Postgres fixture for Stage 0
新建 backend/tests/fixtures/postgres.py,提供两层 fixture:
  - postgres_container (session 级):用 PostgresContainer 起 postgres:16-alpine
  - postgres_url (function 级):每 test 一个 ephemeral DB,teardown 时
    pg_terminate_backend 清掉残连后 DROP DATABASE

为什么 per-DB 而不是 per-schema:asyncpg + SQLAlchemy 不通过 URL 传 search_path,
per-DB 一次 ~50ms 开销可接受,让 test 代码不感知 schema。

Docker 不可用时 fixture 自动 pytest.skip 而非 error,dev 环境无 Docker
仍能跑其余 3086 个测试。

注册 pytest mark `postgres`、把 tests/ 加 sys.path 让 pytest_plugins
按 `fixtures.postgres` 路径解析(不加 tests/__init__.py 避免干扰
现有 pytest 发现行为)。

Stage 0 PR1 T1.3.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-10 22:44:02 +08:00
1445043649 ae4ea46be2 feat(docker): add postgres service to dev compose
新增 postgres:16-alpine service + named volume + healthcheck,作为 Stage 0+
默认 DB backend。生产可通过 DATABASE_URL 指向远程 RDS 时跳过此 service。
端口 / 用户 / 密码 / 库名都走 env var 覆盖(POSTGRES_USER/PORT/...
默认 deerflow/5432/deerflow_dev)。

YAML 通过 `docker compose config` 校验;live healthcheck 待 docker daemon 起后由
T1.3 testcontainers 路径覆盖。

Stage 0 PR1 T1.2.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-10 22:40:12 +08:00
1445043649 fab85b14a6 build(deps): add postgres-test optional extra for testcontainers
新增 postgres-test extra(harness + backend),引入 testcontainers[postgres]
4.14.2 用于 Stage 0 PR1 的 PG fixture。复用现有 postgres extra 的
asyncpg/psycopg/langgraph-checkpoint-postgres,pytest-asyncio 沿用 dev group。

Stage 0 PR1 T1.1.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-10 22:38:50 +08:00
1445043649 a74b88a45c docs(superpowers): add Stage 0 multi-tenant foundation master plan
8 PR 串成的 stage-level 拆解(Postgres 切换 + workspaces schema + auth
扩字段 + 入口路由强校验 + CI boundary + headless schema 预留)。每个 PR
含 commit-sized TDD task list 和验收清单。执行时各 PR 独立 review
checkpoint。

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-10 22:33:25 +08:00
1445043649 89fe54cc07 docs(multi-tenant): 加入汇总索引、跨文档一致性修订、Postgres 切换前移到 Stage 0
* 新增 README.zh-CN.md 汇总索引:ADR 状态表 + Stage 0-4 业务目标 / 技术路径 /
  验证方式 + Stage↔ADR 对照矩阵 + 不可逆决策一览 + 用语映射 + FAQ + 阅读路径
* 新增 workspace-schema-design.zh-CN.md(Stage 0 schema 锁定版)一并入库
* 7 份 ADR 顶部加"代码命名"映射行(tenant_id ↔ workspace_id)
* ADR-002 §1 加分期落地提示,明确 K8s 推迟到 Stage 3
* ADR-005 §5"第 1/2 阶段"补出与 rollout Stage 2/3 的映射
* ADR-007 §4 加 /api/v1/ 反向链接;§8 加 tid → wid 字段名映射
* headless-api §0/§7 把"SaaS + on-prem 双主线"改为"SaaS 主线、schema 兼容 on-prem"
* phased-rollout 去除重复的"Go/No-Go 进入 Stage 2"段

Postgres 切换从 Stage 1 提前到 Stage 0:Stage 0 已要 ALTER 4 张表加
workspace_id,先 SQLite 再 PG 是纯返工;Stage 0 没有生产数据,迁移阻力最小。
同步调整 phased-rollout / headless-api / phase-0-plan / workspace-schema-design /
README 中的时间盒(Stage 0: 3-4→4-5 周;Stage 1: 10-15→8-13 周)、不可逆决策
清单、PR 顺序、轨道前置依赖。

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-10 21:39:58 +08:00
1445043649 9ff790554d docs(persistence): 给 ORM 表/字段补充中文注释
通过 SQLAlchemy 的 comment= 给 5 张持久化表(users / threads_meta / runs /
run_events / feedback)的所有字段以及表本身加上中文注释,便于读代码、
生成文档与未来切到 Postgres 时直接落库为 COMMENT ON。

SQLite 引擎本身不支持 COMMENT ON,运行时不会改变 .schema 输出。

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-10 18:50:17 +08:00
1445043649 ecc9339ede docs(multi-tenant): 加入 Headless API 改造轨道(Pattern A/B)+ Stage 1 双轨修订
新增 headless-api-track 文档,把 DeerFlow 作为后端服务的两种集成
pattern 全部纳入设计:

- Pattern A(业务 backend 代理):业务系统 backend 用 API key 调,
  浏览器走他们的 backend。覆盖 IM channels + server-to-server 集成。
- Pattern B(自研 web 浏览器直连):业务 backend 调
  /api/v1/auth/exchange-token 换 5-15 min 短期 JWT → 浏览器拿 JWT 直连
  含 SSE。核心 4 件事:exchange-token endpoint、ServiceTokenAuthBackend
  (AuthMiddleware 第三条路径)、workspaces.allowed_origins + CORS
  中间件、SSE 跨域验证。
- 不做 widget / iframe(纯 API)。

数据模型:service_accounts / api_keys / external_users 表(schema 在
Stage 0 末加上不阻塞);身份模式三态(collapsed / external_passthrough
/ both)按 endpoint 分支。

Stage 1 改为三轨并行:付费 SaaS / Pattern A / Pattern B;时间盒从
6-10 周延到 10-15 周。Pattern B 依赖 Pattern A 完成,建议 Stage 1
末 1-2 周做。

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-09 23:29:46 +08:00
1445043649 fd8d0d637d docs(multi-tenant): 加入 Stage 0 代码地图
把 Stage 0 必做项落到具体文件 + 行号,覆盖 7 个子系统:
auth、user 数据模型、thread 入口路由、ThreadDataMiddleware + 路径系统、
仓储访问模式、setup / 注册流程、CSRF + ContextVar 注入。

文档结构:每节有【关键文件 + 行号】、【关键函数 / 类】、【当前数据流】、
【Stage 0 改动锚点】;外加改动影响面总览(每条 Stage 0 必做项映射到
具体代码位置)、不可逆决策落点、推荐阅读顺序(11 个文件,1.5-2.5h
粗读)。

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-09 23:29:27 +08:00
1445043649 35ae97d0a6 docs(multi-tenant): Stage 2 加入 per-user skill 覆盖与 config 设计
补充团队 workspace 内不同成员的 skill 区分模型:

- 启用状态:workspace 级默认 + per-user 覆盖
  (final_enabled = user_override ?? workspace_default)
- skill 私有配置(API key 等):per-user 强制隔离 + KMS 加密
- 上传权限:仅 owner / admin,成员只能 enable/disable + 填自己 config

PR 顺序新增第 8 步,依赖 RBAC + KMS 已就位。

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-09 22:47:05 +08:00
1445043649 e78ed687a6 docs(multi-tenant): 加入按规模分期的落地方案(02-rollout)
基于已收敛的目标客户画像(个人用户为主 + 小团队 / 中心化 SaaS /
freemium)把 7 份 ADR 拆成 5 期落地:

- Stage 0:workspace 模型立起来 + auth 收紧(不可逆决策集中在此)
- Stage 1:Postgres + Quota 必落(freemium 不上 quota = 信用卡递给攻击者)
- Stage 2:DeerFlow 表 RLS、KMS、ObjectStorage S3、付费分层
- Stage 3:K8s sandbox + BYO key + audit DB 拆分
- Stage 4:SSO / custom domain / per-tenant DB(按 enterprise 客户合同驱动)

关键取舍:① Stage 0 末就把 workspace_id 列加到 SQLite,避免 Stage 1
切 Postgres 时再补;② Quota 比 ADR-003 原稿提前一档到 Stage 1;③ K8s
sandbox 推到 Stage 3,AioSandbox + 出网白名单 + 资源限额撑到几千用户。

时间盒:最小可付费 4 个月(Stage 0+1)/ 风险可控增长 8 个月(+Stage 2)
/ 全功能 14 个月(+Stage 3)。

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-09 22:28:55 +08:00
1445043649 8bdd308ea0 docs(multi-tenant): 据审计对齐 ADR-002/003/004/005 与 phase-0 计划
非 spike 驱动的对齐改动:
- ADR-002/004/005:头部加现状提示 + 链接审计报告(沙箱出网/资源缺位、
  token_version 已存在 MembershipCache 全新建、ObjectStorage/7 表/KMS
  全部从 0 起)
- ADR-003 LLM 计费:修正 TokenUsageMiddleware 当前只 log 不持久化的描述;
  补充 create_chat_model sync→async 改造的连带影响说明
- phase-0 计划:新增 §3.5 底座先行(Postgres 测试夹具 / ObjectStorage
  Protocol / KMS 抽象 3 件并行做),时间盒 2 → 3 周;ADR-006/007 摘要
  对齐;DoD 加底座骨架检查项

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-09 21:03:32 +08:00
1445043649 27c4f14233 docs(multi-tenant): 加入 ADR 审计 + spike,并据其修订 ADR-001/006/007
新增两份评审产出物:
- adr-vs-code-audit:7 份 ADR 与现状代码的差异核对,标注每条假设是否成立
- adr-spike-langgraph-postgres:实测 langgraph-checkpoint-postgres==3.0.5
  注入能力,确认不存在 connection_factory 参数,且 psycopg_pool 自带的
  configure callback 不是 per-acquire hook

据 spike 与审计修订三份 ADR:
- ADR-001 数据隔离:LangGraph 表改为应用层强校验 + threads_meta unique
  约束兜底(不再挂 RLS、不 ALTER 表);hook 点从 AssistantsCompat 修正
  为 threads.py + thread_runs.py
- ADR-006 运行时与渠道:§2.1 完全重写为应用层强校验;MCP OAuth token
  从"无持久化进程内存"直接做加密 DB;channel store binding 改为新建
  channel_bindings 表
- ADR-007 路由与前端:删除 Better Auth 假设(前端实际无此依赖),改为
  扩展现有 auth/jwt.py TokenPayload 加 tid/role 字段

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-09 21:03:16 +08:00
1445043649 dce5e9598b docs(multi-tenant): add ADRs and phase-0 plan for multi-tenant redesign
Add 7 ADRs and a phase-0 plan covering the multi-tenant redesign of
DeerFlow, plus an architecture-overview snapshot of the current state.

ADRs:
- 001 data isolation: row-level tenant_id + Postgres RLS, including
  LangGraph-owned checkpoint tables (subquery RLS or column upgrade path).
- 002 sandbox isolation: K8s namespace + gVisor + NetworkPolicy default-
  deny, threat model and pod spec defaults.
- 003 LLM key & billing: hybrid platform/BYO with pessimistic reservation
  to handle the "ghost token" overflow on the last call, plus a usage
  category split for memory/title/summarization charges.
- 004 RBAC: two-level (owner/admin/member), JWT-with-role + 30s LRU
  cache for reads, strict DB lookup for sensitive writes, token_version
  bump as the single revocation path.
- 005 storage topology: Postgres (structured) + S3-compatible object
  store (large objects) + emptyDir (ephemeral); explicit treatment of
  the extensions_config.json migration's downstream effects.
- 006 runtime & channel tenancy: per-tenant MCP cache, dual-track skills
  loader, sandbox provider routing by namespace, internal LLM call
  billing, IM channel-to-tenant binding model.
- 007 routing & frontend: path-slug URL form, JWT-only API auth,
  TenantProvider, hard-reload tenant switch, Better Auth integration.

These docs are decision records; no code changes are included.
2026-05-08 23:45:33 +08:00
DanielWalnut 2b1fcb3e43 fix(task): remove max_turns parameter from task tool interface (#2783)
Unit Tests / backend-unit-tests (push) Has been cancelled
Frontend Unit Tests / frontend-unit-tests (push) Has been cancelled
Lint Check / lint (push) Has been cancelled
Lint Check / lint-frontend (push) Has been cancelled
* fix(task): remove max_turns parameter from task tool interface

Subagents should always use their configured max_turns value. Exposing
this parameter allowed callers to override the admin-configured limit,
which is undesirable. The value is now exclusively driven by subagent
config (per-agent overrides and global defaults in config.yaml).

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* Potential fix for pull request finding

Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>

---------

Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
Co-authored-by: Willem Jiang <willem.jiang@gmail.com>
Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
2026-05-08 15:05:24 +08:00
He Wang 7de9b5828b fix(tools): introduce Runtime type alias to eliminate Pydantic serialization warning (#2774)
* fix(tools): introduce Runtime type alias to eliminate Pydantic serialization warning

Add deerflow/tools/types.py with:

    Runtime = ToolRuntime[dict[str, Any], ThreadState]

Replace every runtime: ToolRuntime[ContextT, ThreadState] and
runtime: ToolRuntime[dict[str, Any], ThreadState] annotation in
sandbox/tools.py, present_file_tool.py, task_tool.py, view_image_tool.py,
and skill_manage_tool.py with the new Runtime alias.

The unbound ContextT TypeVar (default None) caused
PydanticSerializationUnexpectedValue warnings on every tool call because
LangChain's BaseTool._parse_input calls model_dump() on the auto-generated
args_schema while DeerFlow passes a dict as runtime context.
Binding the context to dict[str, Any] aligns Pydantic's serialization
expectations with reality and removes the noise from all run modes.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Co-authored-by: Cursor <cursoragent@cursor.com>

* fix(tools): extend Runtime alias to setup_agent and update_agent tools

Replace bare ToolRuntime annotations in setup_agent_tool.py and
update_agent_tool.py with the shared Runtime alias introduced in the
previous commit, and add both tools to the Pydantic serialization
warning regression test (13 cases total).

Co-authored-by: Cursor <cursoragent@cursor.com>

* test(tools): loosen Pydantic warning filter to avoid version-specific format

Replace the brittle "field_name='context'" substring check with a looser
"context" match so the assertion stays valid if Pydantic changes its
internal warning format across versions.

Co-authored-by: Cursor <cursoragent@cursor.com>

* test(tools): simplify warning filter and clean up docstring

Remove the "context" substring condition from the Pydantic warning
filter — asserting that no PydanticSerializationUnexpectedValue fires
at all is both simpler and more comprehensive, since the test payload
contains only the tool's own args plus runtime.

Also update the module docstring to remove the version-specific warning
format example that was inconsistent with the looser filter.

Co-authored-by: Cursor <cursoragent@cursor.com>

---------

Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-05-08 14:50:33 +08:00
Eilen Shin 37db689349 fix(events): serialize structured db event content (#2762) 2026-05-08 10:17:17 +08:00
Eilen Shin bd45cb2846 fix(sandbox): disable msys path conversion (#2766) 2026-05-08 10:13:11 +08:00
Eilen Shin 5fd0e6ac89 fix(middleware): sync raw tool call metadata (#2757) 2026-05-08 10:08:53 +08:00