惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

罗磊的独立博客
爱范儿
爱范儿
OSCHINA 社区最新新闻
OSCHINA 社区最新新闻
让小产品的独立变现更简单 - ezindie.com
让小产品的独立变现更简单 - ezindie.com
博客园_首页
博客园 - 叶小钗
酷 壳 – CoolShell
酷 壳 – CoolShell
Apple Machine Learning Research
Apple Machine Learning Research
云风的 BLOG
云风的 BLOG
量子位
博客园 - 三生石上(FineUI控件)
Stack Overflow Blog
Stack Overflow Blog
小众软件
小众软件
奇客Solidot–传递最新科技情报
奇客Solidot–传递最新科技情报
V
V2EX
人人都是产品经理
人人都是产品经理
V
Visual Studio Blog
Jina AI
Jina AI
L
LangChain Blog
M
MIT News - Artificial intelligence
MongoDB | Blog
MongoDB | Blog
Last Week in AI
Last Week in AI
Martin Fowler
Martin Fowler
WordPress大学
WordPress大学

Recent Commits to openclaw:main

test: merge chat side-result checks · openclaw/openclaw@ddd2c2a test: merge cron history checks · openclaw/openclaw@f7eb746 test: merge responsive navigation shell checks · openclaw/openclaw@c2e4b47 docs(changelog): add codex oauth fixes · openclaw/openclaw@628e6cd test: merge navigation routing cases · openclaw/openclaw@5d8cecb Tests: mock channel registry bundled fallback · openclaw/openclaw@2b08233 Secrets: avoid broad web search discovery for single plugin config · openclaw/openclaw@a464f59 test: merge config view browser checks · openclaw/openclaw@20cf511 fix(status): align oauth health with runtime · openclaw/openclaw@eed7116 feat: add macOS screen snapshots for monitor preview (#67954) thanks … · openclaw/openclaw@f377db1 fix: report shared auth scopes in hello-ok (#67810) thanks @BunsDev · openclaw/openclaw@0b6c39b Auto-reply: avoid eager bundled route fallback · openclaw/openclaw@3ea1bf4 Tests: narrow session binding contract setup · openclaw/openclaw@54e4e16 fix(macOS): enable undo/redo in webchat composer text input (#34962) · openclaw/openclaw@00951dc Tests: speed up channel setup promotion · openclaw/openclaw@82b529a Docs: refresh agent instructions · openclaw/openclaw@5775fe2 fix(auth): serialize OAuth refresh across agents to fix #26322 (#67876) · openclaw/openclaw@8e79080 test: allow ollama public surface boundary test · openclaw/openclaw@7d4f1a6 Docs: add test performance guardrails · openclaw/openclaw@89706d3 Tests: restore context-engine usage proof · openclaw/openclaw@e4c4f95 Tests: slim context engine runtime coverage · openclaw/openclaw@74c198f ci: retry failed custom checkouts · openclaw/openclaw@0ee5baf test: trim duplicate provider auth onboarding cases · openclaw/openclaw@1ffc02e matrix: fix sessions_spawn --thread subagent session spawning (#67643) · openclaw/openclaw@1ce2596 test: reduce auth choice fixture churn · openclaw/openclaw@857b9cd test: mock health status config boundaries · openclaw/openclaw@9d5ab4a test: mock onboard config io boundary · openclaw/openclaw@299694d test: mock legacy state plugin boundaries · openclaw/openclaw@2713089 test: mock channel install boundaries · openclaw/openclaw@b945248 test: mock doctor preview channel boundaries · openclaw/openclaw@b1a3ad4
docs: add Mantis Slack desktop runbook · openclaw/opencla...
steipete · 2026-05-06 · via Recent Commits to openclaw:main

@@ -0,0 +1,202 @@

1+

---

2+

summary: "Operator runbook for Mantis Slack desktop QA: GitHub dispatch, local CLI, warm VNC leases, hydrate modes, timing interpretation, artifacts, and failure handling."

3+

read_when:

4+

- Running Mantis Slack desktop QA from GitHub or locally

5+

- Debugging slow Mantis Slack desktop runs

6+

- Choosing source, prehydrated, or warm-lease mode

7+

- Posting screenshot and video evidence to a PR

8+

title: "Mantis Slack Desktop Runbook"

9+

---

10+11+

Mantis Slack desktop QA is the real-UI lane for Slack-class bugs that need a

12+

Linux desktop, VNC rescue, Slack Web, a real OpenClaw gateway, screenshots,

13+

videos, and a PR evidence comment.

14+15+

Use it when unit tests or the headless Slack live lane cannot prove the bug.

16+17+

## Storage Model

18+19+

Mantis uses three different storage layers:

20+21+

- Provider image: owned by Crabbox and stored in the cloud provider account.

22+

It contains machine capabilities such as Chrome/Chromium, ffmpeg, scrot,

23+

Node/corepack/pnpm, native build tools, and empty cache directories.

24+

- Warm lease state: owned by the current operator session. It can contain a

25+

logged-in browser profile, `/var/cache/crabbox/pnpm`, and a prepared source

26+

checkout while the lease is alive.

27+

- Mantis artifacts: owned by the OpenClaw run. They live under

28+

`.artifacts/qa-e2e/mantis/...`, then GitHub Actions uploads them and the

29+

Mantis GitHub App comments inline evidence on the PR.

30+31+

Never put secrets, browser cookies, Slack login state, repository checkouts,

32+

`node_modules`, or `dist/` into a prebaked provider image.

33+34+

## GitHub Dispatch

35+36+

Run the workflow from `main`:

37+38+

```bash

39+

gh workflow run mantis-slack-desktop-smoke.yml \

40+

--ref main \

41+

-f candidate_ref=<trusted-ref-or-sha> \

42+

-f pr_number=<pr-number> \

43+

-f scenario_id=slack-canary \

44+

-f crabbox_provider=aws \

45+

-f keep_vm=false \

46+

-f hydrate_mode=source

47+

```

48+49+

Allowed `candidate_ref` values are intentionally narrow because the workflow

50+

uses live credentials: current `main` ancestry, release tags, or an open PR head

51+

from `openclaw/openclaw`.

52+53+

The workflow writes:

54+55+

- uploaded artifact: `mantis-slack-desktop-smoke-<run-id>-<attempt>`;

56+

- inline PR comment from the Mantis GitHub App;

57+

- `slack-desktop-smoke.png`;

58+

- `slack-desktop-smoke.mp4`;

59+

- `slack-desktop-smoke-preview.gif`;

60+

- `slack-desktop-smoke-change.mp4`;

61+

- `mantis-slack-desktop-smoke-summary.json`;

62+

- `mantis-slack-desktop-smoke-report.md`;

63+

- remote logs such as `slack-desktop-command.log`, `openclaw-gateway.log`,

64+

`chrome.log`, and `ffmpeg.log`.

65+66+

The PR comment is updated in place by the hidden

67+

`<!-- mantis-slack-desktop-smoke -->` marker.

68+69+

## Local CLI

70+71+

Cold source proof:

72+73+

```bash

74+

pnpm openclaw qa mantis slack-desktop-smoke \

75+

--provider aws \

76+

--class standard \

77+

--gateway-setup \

78+

--credential-source convex \

79+

--credential-role maintainer \

80+

--provider-mode live-frontier \

81+

--model openai/gpt-5.4 \

82+

--alt-model openai/gpt-5.4 \

83+

--scenario slack-canary \

84+

--hydrate-mode source

85+

```

86+87+

Keep the VM for VNC rescue:

88+89+

```bash

90+

pnpm openclaw qa mantis slack-desktop-smoke \

91+

--provider aws \

92+

--class standard \

93+

--gateway-setup \

94+

--scenario slack-canary \

95+

--keep-lease

96+

```

97+98+

Open VNC:

99+100+

```bash

101+

crabbox vnc --provider aws --id <cbx_id> --open

102+

```

103+104+

Reuse a warm lease:

105+106+

```bash

107+

pnpm openclaw qa mantis slack-desktop-smoke \

108+

--provider aws \

109+

--lease-id <cbx_id-or-slug> \

110+

--gateway-setup \

111+

--scenario slack-canary \

112+

--hydrate-mode source

113+

```

114+115+

Use `--hydrate-mode prehydrated` only when the reused remote workspace already

116+

has `node_modules` and a built `dist/`. Mantis fails closed if those are

117+

missing.

118+119+

## Hydrate Modes

120+121+

| Mode | Use when | Remote behavior | Tradeoff |

122+

| ------------- | ----------------------------------------- | ------------------------------------------------------------------------------------- | -------------------------------------------------------- |

123+

| `source` | Normal PR proof, cold machines, CI | Runs `pnpm install --frozen-lockfile --prefer-offline` and `pnpm build` inside the VM | Slowest, strongest source-checkout proof |

124+

| `prehydrated` | You intentionally prepared a reused lease | Requires existing `node_modules` and `dist/`; skips install/build | Fast, but only valid for operator-controlled warm leases |

125+126+

GitHub Actions always prepares the candidate checkout before the VM run. Its

127+

pnpm store is cached by OS, Node version, and lockfile. The VM source run also

128+

uses `/var/cache/crabbox/pnpm` when present.

129+130+

## Timing Interpretation

131+132+

`mantis-slack-desktop-smoke-report.md` includes phase timings:

133+134+

- `crabbox.warmup`: cloud provider boot, desktop/browser readiness, and SSH.

135+

- `crabbox.inspect`: lease metadata lookup.

136+

- `credentials.prepare`: Convex credential lease acquisition.

137+

- `crabbox.remote_run`: sync, browser launch, OpenClaw install/build or

138+

hydrate validation, gateway startup, screenshot, and video capture.

139+

- `artifacts.copy`: rsync back from the VM.

140+141+

`crabbox.remote_run` can be marked `accepted` when Crabbox returns a non-zero

142+

remote status after Mantis has copied metadata proving that the OpenClaw gateway

143+

is alive and the setup completed. Treat `accepted` as pass-with-explanation,

144+

not a failed scenario.

145+146+

If the run is slow:

147+148+

- warmup dominates: prebake or promote a better Crabbox provider image;

149+

- remote_run dominates in `source`: use a warm lease, improve pnpm store reuse,

150+

or move machine prerequisites into the provider image;

151+

- remote_run dominates in `prehydrated`: the remote workspace was not actually

152+

ready, or the gateway/browser/Slack setup is slow;

153+

- artifact copy dominates: inspect video size and artifact directory contents.

154+155+

## Evidence Checklist

156+157+

A good PR comment should show:

158+159+

- scenario id and candidate SHA;

160+

- GitHub Actions run URL;

161+

- artifact URL;

162+

- inline screenshot;

163+

- inline animated preview when available;

164+

- full MP4 and trimmed MP4 links;

165+

- pass/fail status;

166+

- timing summary in the attached report.

167+168+

Do not commit screenshots or videos into the repository. Keep them in GitHub

169+

Actions artifacts or the PR comment.

170+171+

## Failure Handling

172+173+

If the workflow fails before the VM run, inspect the Actions job first. Typical

174+

causes are untrusted `candidate_ref`, missing environment secrets, or candidate

175+

install/build failure.

176+177+

If the VM run fails but screenshots were copied back, inspect:

178+179+

```bash

180+

cat mantis-slack-desktop-smoke-report.md

181+

cat mantis-slack-desktop-smoke-summary.json

182+

cat slack-desktop-command.log

183+

cat openclaw-gateway.log

184+

cat chrome.log

185+

cat ffmpeg.log

186+

```

187+188+

If the run kept the lease, open VNC with the report's `crabbox vnc ...` command.

189+

Stop the lease when done:

190+191+

```bash

192+

crabbox stop --provider aws <cbx_id-or-slug>

193+

```

194+195+

If Slack login expired, repair it in VNC on a kept lease and rerun with

196+

`--lease-id`. Do not bake that browser profile into a provider image.

197+198+

Related docs:

199+200+

- [QA overview](qa-e2e-automation.md)

201+

- [Slack channel](../channels/slack.md)

202+

- [Testing](../help/testing.md)