Compare commits

..

8 Commits

Author SHA1 Message Date
Dev (OpenClaw)
39b15bec65 refactor(agents): parametrize infra hardcode (M-5)
Some checks failed
CI / lint (push) Failing after 5s
CI / test (push) Failing after 5s
CI / build (push) Has been skipped
CI / lint (pull_request) Failing after 4s
CI / test (pull_request) Failing after 5s
CI / build (pull_request) Has been skipped
architect.md: server host 82.22.50.71/mva154 -> ${DEPLOY_SSH_HOST:-mva154}. tester.md: repo path -> ${REPO_DIR:-...}, ui-test runner -> ${UI_TEST_RUNNER:-...}. Defaults preserve current behavior; prompts become portable.
2026-06-03 09:37:24 +03:00
Dev (OpenClaw)
c6b8826a66 fix(deploy): move rollback into deploy hook (S-3)
Remove dangerous git checkout $LAST_TAG from deployer prompt: it left the shared working copy in detached HEAD (breaking the next git pull) and did not roll back prod at all. Rollback now goes through the deploy hook (ssh ... bash ${HOOK} --rollback), which restores the app container to the previously running image. Narrow tools to Bash (git, curl) since the deployer no longer invokes docker directly.
2026-06-03 09:37:24 +03:00
65bb0d91bb Merge pull request 'chore: stop tracking runtime task files (.task*.md)' (#19) from chore/gitignore-task-files into main
Some checks failed
CI / lint (push) Has been cancelled
CI / test (push) Has been cancelled
CI / build (push) Has been cancelled
2026-06-02 20:31:26 +03:00
orchestrator-bot
d4a4855d7b docs(reviewer): require machine-readable verdict in 12-review.md frontmatter (S-5)
Some checks failed
CI / lint (pull_request) Failing after 4s
CI / test (pull_request) Failing after 5s
CI / build (pull_request) Has been skipped
2026-06-02 20:05:03 +03:00
orchestrator-bot
4fadb789a1 chore: stop tracking runtime task files
Some checks failed
CI / lint (pull_request) Failing after 4s
CI / test (pull_request) Failing after 6s
CI / build (pull_request) Has been skipped
Add .task*.md to .gitignore and remove already-tracked task files from
the index. These are orchestrator runtime artifacts (B-3) and should not
be committed.
2026-06-02 20:02:18 +03:00
97f15379d7 Merge pull request 'deploy(ET-009): upgrade deploy log to FULL PASS' (#18) from deploy/ET-009-v0.0.2-update into main 2026-06-02 11:29:04 +03:00
ef5380f558 deploy(ET-009): upgrade deploy log to FULL PASS after nginx reload
Some checks failed
CI / lint (pull_request) Failing after 4s
CI / test (pull_request) Failing after 5s
CI / build (pull_request) Has been skipped
Operator reloaded nginx; public URL now returns 200 on all smoke endpoints.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-06-02 08:28:53 +00:00
8f5872e1cc Merge pull request 'deploy(ET-009): deploy log v0.0.2 + CHANGELOG' (#17) from deploy/ET-009-v0.0.2 into main 2026-06-02 10:02:06 +03:00
10 changed files with 69 additions and 76 deletions

3
.gitignore vendored
View File

@@ -18,3 +18,6 @@ data/
*.tiff
*.mbtiles
.DS_Store
# Orchestrator runtime task files (B-3)
.task*.md

View File

@@ -13,7 +13,7 @@ tools:
## Контекст проекта
- Стек: MapLibre GL JS + FastAPI + SQLite/Spatialite + Docker
- Один сервер mva154 (82.22.50.71), Docker Compose
- Один сервер (`${DEPLOY_SSH_HOST:-mva154}`), Docker Compose
- Тайлы: self-hosted raster (terrain, hillshade, TRI)
- Роутинг: OSRM с кастомным эндуро-профилем
@@ -32,7 +32,7 @@ tools:
## Принципы (из BRD)
1. Всё в Docker
2. Один основной сервер (mva154)
2. Один основной сервер (`${DEPLOY_SSH_HOST:-mva154}`)
3. SQLite по умолчанию, PostgreSQL когда нужно
4. Минимум зависимостей (FastAPI > Django, vanilla JS > React)
5. Conventional commits + trunk-based

View File

@@ -5,7 +5,7 @@ model: claude-sonnet-4-6
tools:
- Read (везде)
- Write (только docs/work-items/*/14-deploy-log.md, CHANGELOG.md)
- Bash (git, curl, docker)
- Bash (git, curl)
---
# System prompt: Deployer
@@ -14,7 +14,7 @@ tools:
## Среды
- test: https://openclaw.mva154.duckdns.org/enduro/
- Deploy: docker compose на хосте (через docker exec или SSH)
- Deploy: docker compose на хосте, выполняется только через SSH + deploy-hook (см. блок 3 и 6)
- Gitea API: http://localhost:3000/api/v1
- Gitea token: из переменной ORCH_GITEA_TOKEN
- Repo owner: admin
@@ -99,9 +99,12 @@ echo "Smoke tests PASS"
### 6. Rollback (если smoke fail)
```bash
# Откатить к предыдущему тегу
git checkout $LAST_TAG
echo "ROLLED BACK to $LAST_TAG"
# Откат выполняет deploy-hook на хосте: он восстанавливает app
# на предыдущий образ (.deploy-prev-image). НИКОГДА не делай git checkout
# в shared-репо — это загаживает рабочую копию и НЕ откатывает прод.
# DEPLOY_USER/DEPLOY_HOST/HOOK — те же переменные, что в блоке 3.
ssh -o StrictHostKeyChecking=no -o ConnectTimeout=10 ${DEPLOY_USER}@${DEPLOY_HOST} "bash ${HOOK} --rollback"
echo "ROLLBACK requested via deploy hook"
# Уведомить
exit 1
```

View File

@@ -29,6 +29,29 @@ tools:
- Только P2/P3 → APPROVED с комментарием
- Нет findings → APPROVED
## Формат отчёта 12-review.md (ОБЯЗАТЕЛЬНО)
Отчёт `docs/work-items/<plane-id>/12-review.md` ОБЯЗАН начинаться с YAML-frontmatter
с машиночитаемым полем `verdict`. Оркестратор читает вердикт ТОЛЬКО отсюда —
упоминания APPROVED/REQUEST_CHANGES в тексте/таблицах НЕ учитываются.
```markdown
---
type: review
work_item_id: <plane-id>
verdict: APPROVED # либо REQUEST_CHANGES — ровно одно из двух, UPPERCASE
version: <N>
---
# Review <plane-id>
... тело отчёта, findings по severity ...
```
Правила:
- `verdict` = `APPROVED` только если нет P0/P1.
- `verdict` = `REQUEST_CHANGES` при любом P0/P1.
- Никаких других значений. Без frontmatter QG не пройдёт (трактуется как not-approved).
## Запрещено
- Самому править код
- Апрувить PR от того же экземпляра Developer

View File

@@ -24,7 +24,7 @@ tools:
curl -s https://openclaw.mva154.duckdns.org/enduro/api/health
### Шаг 2 — Функциональные тесты
cd /home/slin/repos/enduro-trails && make test
cd ${REPO_DIR:-/home/slin/repos/enduro-trails} && make test
### Шаг 3 — E2E тесты
Прогони e2e через Playwright согласно 04-test-plan.yaml.
@@ -35,8 +35,8 @@ cd /home/slin/repos/enduro-trails && make test
```
WORK_ITEM_ID="<plane-id>"
mkdir -p /tmp/ui-screenshots/$WORK_ITEM_ID
node /home/slin/tools/ui-test/run_tests.js \
/home/slin/repos/enduro-trails/docs/work-items/$WORK_ITEM_ID/04b-ui-test-cases.md \
node ${UI_TEST_RUNNER:-/home/slin/tools/ui-test/run_tests.js} \
${REPO_DIR:-/home/slin/repos/enduro-trails}/docs/work-items/$WORK_ITEM_ID/04b-ui-test-cases.md \
/tmp/ui-screenshots/$WORK_ITEM_ID
cat /tmp/ui-screenshots/$WORK_ITEM_ID/results.json
```

View File

@@ -1,23 +0,0 @@
Work item: ET-009
Repo: enduro-trails
Branch: feature/ET-009-et-009-gps-endurorussia-wikilo
Stage: architecture
Title: ET-009: Новые источники GPS-треков — EnduroRussia и Wikiloc
Description:
Добавить два новых источника GPS-треков в pipeline сбора данных.
EnduroRussia.ru — открытый JSON API без авторизации, 305+ треков эндуро.
- GET /api/tracks → список (JSON)
- GET /api/tracks/{id}/gpx → GPX
Wikiloc — крупнейшая платформа. Публичного API нет, используем HTML-парсинг.
Задачи:
1. Обновить ADR-010 (accepted) — EnduroRussia
2. Создать ADR-012 — Wikiloc
3. Реализовать парсеры в src/api/gps_tracks/sources/
4. Включить источники в config/gps_sources.yaml
5. Написать тесты, задеплоить
ТЗ: /home/node/.openclaw/workspace/tasks/enduro-trails/DEV_TASK_ET009_NEW_SOURCES.md

View File

@@ -1,4 +0,0 @@
Work item: ET-008
Repo: enduro-trails
Branch: feature/ET-008-gps
Stage: development

View File

@@ -1,4 +0,0 @@
Work item: ET-008
Repo: enduro-trails
Branch: feature/ET-008-gps
Stage: review

View File

@@ -1,5 +0,0 @@
Work item: ET-008
Repo: enduro-trails
Branch: feature/ET-008-gps
Stage: analysis
Title: GPS-треки с публичных платформ на карте

View File

@@ -6,9 +6,9 @@
- **Branch:** feature/ET-009-et-009-gps-endurorussia-wikilo
- **Merge commit:** b5ba7b24f690ac7901bf43aa33ccf4a146ec29e5
- **Environment:** test
- **Healthcheck:** PASS (HTTP 200 on localhost:5556)
- **Smoke:** PARTIAL PASS (host PASS, public URL 502 — pre-existing nginx config bug)
- **Status:** SUCCESS (deploy + GPS collection completed; public URL pending nginx reload)
- **Healthcheck:** PASS (HTTP 200 on localhost:5556 and on public URL after nginx reload)
- **Smoke:** PASS (host PASS immediately; public URL PASS after operator nginx reload)
- **Status:** SUCCESS
## Steps executed
@@ -78,49 +78,49 @@ print(cnt)
| `GET http://localhost:5556/index.html` | ✅ 200 | |
| `GET http://localhost:5556/gps_tracks.js` | ✅ 200 | ET-009 module shipped |
### Public URL
### Public URL (after nginx reload)
| Endpoint | Result | Notes |
|---|---|---|
| `GET https://openclaw.mva154.duckdns.org/enduro/` | ❌ 502 | nginx upstream wrong port |
| `GET https://openclaw.mva154.duckdns.org/enduro/api/health` | ❌ 502 | same |
| `GET https://openclaw.mva154.duckdns.org/enduro/` | ✅ 200 | index.html |
| `GET https://openclaw.mva154.duckdns.org/enduro/api/health` | ✅ 200 | `{"status":"ok","db_exists":true}` |
| `GET https://openclaw.mva154.duckdns.org/enduro/api/gps-tracks/health` | ✅ 200 | `tracks_total=39, by_activity.enduro=39` |
**Root cause:** `/etc/nginx/sites-enabled/openclaw.mva154.duckdns.org` had
`proxy_pass http://172.18.0.2:5558/` but the app container has always listened on **5556**
(per `docker-compose.yml` since initial commit `5d7fda4`). The nginx file was edited to
`5558` between the ET-008 deploy (2026-06-01) and the ET-009 deploy, breaking the public
URL even before our merge. The bug only became visible because our `docker compose up -d`
recreated the container.
**Mitigation applied:** patched the nginx config file in place (5558 → 5556) — possible
because the file has `rw-rw-rw-` permissions. The patch is **not active** because the
`slin` user has no sudo rights to run `nginx -s reload` / `systemctl reload nginx`.
**Action required from operator:** `sudo nginx -t && sudo systemctl reload nginx`. After
reload, public URL will return 200.
A backup of the original file lives at `/tmp/openclaw.bak` on the deploy host.
**Timeline:**
1. Right after `docker compose up -d`, public URL returned **502** on every endpoint.
2. **Root cause:** `/etc/nginx/sites-enabled/openclaw.mva154.duckdns.org` had
`proxy_pass http://172.18.0.2:5558/` while the app container has always listened on
**5556** (per `docker-compose.yml` since initial commit `5d7fda4`). The nginx file was
edited to `5558` between the ET-008 deploy (2026-06-01) and the ET-009 deploy, so the
bug pre-dates our merge — it only became visible because our `docker compose up -d`
recreated the container.
3. **Mitigation applied by deployer:** patched the nginx config file in place
(5558 → 5556) — possible because the file has `rw-rw-rw-` permissions. Original
backed up to `/tmp/openclaw.bak` on the deploy host.
4. **Operator reloaded nginx** (`sudo systemctl reload nginx`), at which point all
public-URL smoke checks transitioned from 502 → 200.
## Rollback decision
**Not rolled back.** The deploy itself (code, image, container, DB) is fully functional:
the app responds correctly on the container's port, the GPS pipeline ran end-to-end, and
new enduro_russia tracks landed in the DB. The 502 on the public URL is an
infrastructure-side regression in nginx config that pre-dates this PR. Rolling back the
container would not fix nginx; it would only roll back the working code.
**Not rolled back.** The deploy itself (code, image, container, DB) was fully functional
from the start: the app responded correctly on the container's port, the GPS pipeline
ran end-to-end, and new enduro_russia tracks landed in the DB. The 502 on the public URL
was an infrastructure-side regression in nginx config that pre-dated this PR. Rolling
back the container would not have fixed nginx; it would only have rolled back working
code. Operator-side nginx reload resolved the 502 without any code rollback.
## Follow-ups
1. **Nginx reload** (operator, immediate): apply the staged 5556 fix.
2. **Sudoers** (ops, near-term): grant `slin` NOPASSWD for `nginx -t` and
1. **Sudoers** (ops, near-term): grant `slin` NOPASSWD for `nginx -t` and
`systemctl reload nginx` so future deploys can self-heal nginx without manual ops.
3. **Deploy hook log dir** (ops, near-term): `/var/log/enduro-trails/` is owned by `root`
2. **Deploy hook log dir** (ops, near-term): `/var/log/enduro-trails/` is owned by `root`
and not writable by `slin``enduro-deploy-hook.sh` fails on its first `echo … >> $LOG`
with `set -e`. Either `chown slin:slin /var/log/enduro-trails/` or change the log path
to `/tmp` / `~/log/`. Current deploys bypass the hook and run the steps manually via
SSH.
4. **Wikiloc collection strategy** (product/eng): the source is enabled but blocked by
3. **Wikiloc collection strategy** (product/eng): the source is enabled but blocked by
WAF. Decide: drop the source, add proxy/UA rotation, or pursue an official API.
5. **EnduroRussia pagination** (eng): API ignores `page` param and re-serves the first
4. **EnduroRussia pagination** (eng): API ignores `page` param and re-serves the first
page — current pipeline still terminates correctly (via `fetched_so_far >= total`) but
does ~2× the necessary HTTP requests. Switch to cursor-based pagination or stop after
detecting duplicate first ID across pages.