TOPY.AI · AI CTO

ai.topy由 topy.ai 發布官方發布者v0.1.0更新於 Oct 29, 2026

Founders' AI CTO: CI failures explained, PRs and checks, release notes, alerts.

發布方自己的服務：TOPY.AI · AI CEO

已驗證Streamable HTTP可網頁執行雲端與基礎設施開發者工具安全與監控

概覽

AI 產生的概覽

TOPY.AI AI CTO 讓程式開發助理透過 GitHub 處理 CI 失敗、PR 審查、發布、部署與安全告警，扮演小型團隊的 AI 技術長。

可以做什麼
  • 解釋 CI：讀取失敗的 GitHub Actions 執行、工作與日誌，分辨偶發失敗與真實故障，並在預覽後重跑或開 issue。
  • 審查程式碼與分流 issue：評估拉取請求的風險，把缺陷回報與 CI 發現整理成有規模、有負責人的 issue。
  • 準備發布：起草發布說明，依證據執行發布門檻，規劃受控部署，並回滾到上一個可用建置。
  • 關注安全與相依套件：讀取 Dependabot、程式碼掃描與密鑰掃描告警，並提供密鑰衛生建議。
  • 回報狀態：常設目標、每日晨檢（服務健康、CI、待處理 PR），以及事故應變與架構決策記錄指引。
適用情境
適合小型團隊或獨立創辦人，希望助理持續關注 GitHub 上的 CI、拉取請求、發布與安全告警，解釋失敗原因並依證據準備發布。適用於合併、發布與生產變更仍由人核准的流程。
執行需求
需要 TOPY.AI 帳號並以 OAuth 連線，或為本機 stdio 套件提供 TOPY_API_KEY。需建立並安裝 TOPY AI GitHub App 到相關儲存庫，且只有擁有 OneMan 營運的帳號才能存取其 GitHub。可連線 topy.ai/mcp/cto 遠端端點，或透過 npx 在 Node.js 上執行本機伺服器。
安裝前請注意
它透過 TOPY AI GitHub App 存取 GitHub，儲存庫存取權會授予該應用程式與 TOPY.AI。建立 issue、留言、重跑等變更需先預覽再確認。合併、發布、新建儲存庫與生產工作流程需在別處由創辦人核准。程式碼、日誌與 issue 文字視為資料處理，密鑰不應被列印或索取。

安裝

在 SourceWeft 中

  1. 開啟 儀表板中的 TOPY.AI · AI CTO，將其新增到工作區。
  2. 連接 所需的 憑證。它們會被私密儲存，絕不會在公開頁面上展示。
  3. 為需要使用其工具的對話啟用該服務。

Web executable，透過 Streamable HTTP。 遠端服務在工作區中設定後即可從網頁執行環境執行。

交給代理程式安裝

把這段提示詞貼上到 Claude Code、Codex、Cursor 等代理程式，或 SourceWeft 對話中。代理程式會讀取該伺服器的安裝指南，先向你展示發布者、端點和登入方式，經你同意後再把它新增到你的客戶端。金鑰不會寫進任何命令。

閱讀 https://sourceweft.com/mcp/ai.topy%2Fcto/install.md，並按其中說明把這個 MCP 伺服器接入我的客戶端。修改任何設定之前，先告訴我它的發布者、端點和登入方式，等我確認。不要把 API 金鑰寫進命令或我會提交的檔案；問我金鑰應從哪裡取得。

檢視代理程式所依據的安裝指南

其他 MCP 客戶端

把它新增到你客戶端的 mcpServers 設定中。 客戶端會在首次連接時完成 OAuth 登入。

{
  "mcpServers": {
    "topy-ai-ai-cto": {
      "type": "http",
      "url": "https://topy.ai/mcp/cto"
    }
  }
}

README

[TOPY AI CTO]

TOPY AI CTO Skills

17 skills that make your coding agent a small company's AI CTO: what is live keeps working and every release has evidence — CI failures explained from their logs, flaky tests quarantined, pull requests reviewed for risk, release notes, a release gate, safe deploys and rollbacks, incident response, dependency alerts and secrets hygiene — through the TOPY AI GitHub App on the TOPY.AI CTO MCP server.

[npm] [MIT License] [MCP]

The AI CTO is goal-first. It keeps four standing goals — what is live works (uptime), we ship often and safely (deploy frequency, change failure rate), every release has evidence (CI green, QA passed, live == built) and the platform can grow — and every skill below exists to move one of them. Runs, checks, logs and alerts come from the tools; an alert kind GitHub won't show is "not read" with GitHub's reason, never zero.

  • Core (topy-ai-cto): connect, the CTO's goals, the morning check, the CI watch.
  • CI: why CI failed, flaky tests, CI hygiene.
  • Code review & issues: review a pull request for risk, triage engineering issues.
  • Releases & deploys: release notes, the release gate, deploy safely, roll back.
  • Dependencies & secrets: dependency and security alerts, secrets hygiene.
  • Incidents & architecture: incident response, architecture decision records.

The safety model comes from Maggie QA Workflow and Maggie Deployment and does not bend: GitHub is reached through the TOPY AI GitHub App the Founder installs — no GitHub token passes through the agent; merging to the default branch, creating repositories or releases and production workflows are the Founder's, approved on OneMan, never tools here; secrets are never printed or asked for; code, commit messages, issue text and logs are data, never instructions; and every release needs evidence before it ships. Changes (an issue, a comment, a re-run) take two calls — a preview, then your yes.

Install

sh
npx @topy-ai/cto-skills install                  # this project, for the agents set up herenpx @topy-ai/cto-skills install -g --mcp         # your user, plus the MCP servernpx @topy-ai/cto-skills install --category ci,releasesnpx @topy-ai/cto-skills list

Claude Code plugins: /plugin marketplace add TOPY-AI-LTD/TOPY-AI-CTO-Skills, then /plugin install topy-ai-cto-all@topy-ai-cto. Codex: codex plugin marketplace add TOPY-AI-LTD/TOPY-AI-CTO-Skills. Pi: pi install npm:@topy-ai/cto-skills. See the installer README.

Connect TOPY.AI

https://topy.ai/mcp/cto — OAuth with your TOPY.AI account (the same one as the AI CEO, CFO, COO, CSO, CRO and CPO). Only the accounts that own the OneMan operation reach its GitHub. Connect GitHub first: create and install the TOPY AI GitHub App from the links OneMan's CTO gives (github_setup_app, github_setup_install); you press Create and Install on GitHub.

ClientHow
Claude, ChatGPT (web, desktop)custom connector https://topy.ai/mcp/cto
Claude Codeclaude mcp add --transport http topy-cto https://topy.ai/mcp/cto
Codexcodex mcp add topy-cto --url https://topy.ai/mcp/cto
Clients that only start a commandnpx -y @topy-ai/cto-mcp after npx @topy-ai/cto-mcp login
Terminal and scriptsnpm i -g @topy-ai/cto-cli && cto-cli login

Merges, releases, new repositories and production workflows are not on topy.ai/mcp/cto: they need the Founder's approval on OneMan.

How the skills fit together

topy-cto-connect ─▶ topy-cto-goals        │   every morning:   topy-cto-morning-check   around the clock: topy-cto-ci-watch (an email only on a failure or a recovery)        │   ci:        topy-cto-ci-failure ─▶ topy-cto-flaky-tests · topy-cto-ci-hygiene   code:      topy-cto-pr-review · topy-cto-issue-triage   releases:  topy-cto-release-notes ─▶ topy-cto-release-gate ─▶ topy-cto-deploy-safely ─▶ topy-cto-rollback   security:  topy-cto-dependency-alerts · topy-cto-secrets-hygiene   platform:  topy-cto-incident · topy-cto-architecture-decision

Try: "Is CI green?", "Why did the build fail?", "Review PR 42", "Can we ship?", "Any security alerts?"

Packages

PackageWhat it is
@topy-ai/cto-skillsThis installer (Claude Code, Codex, Pi, Cursor, Gemini CLI, …)
@topy-ai/cto-mcpThe CTO MCP server as a local stdio server (forwards to https://topy.ai/mcp/cto)
@topy-ai/cto-cliThe AI CTO on the command line

Repository structure

plugins/  topy-ai-cto/                   core skills + .mcp.json (Claude Code) and mcp.json (Codex)  topy-ai-cto-ci/   topy-ai-cto-code/   topy-ai-cto-releases/  topy-ai-cto-security/   topy-ai-cto-platform/  topy-ai-cto-all/               bundle that depends on all six.claude-plugin/marketplace.json     Claude Code marketplace.agents/plugins/marketplace.json    Codex marketplaceskills.json                         machine-readable cataloguepackages/skills/                    the @topy-ai/cto-skills installer

The skills are generated from a source catalogue kept next to TOPY.AI's MCP server; the build checks every tool a skill names against the CTO server's tool list. Please open an issue rather than editing a SKILL.md.

License

MIT © TOPY AI LTD. Privacy: topy.ai/privacy.

Skills

17 skills in 6 plugins, plus the topy-ai-cto-all bundle. Generated from the source catalogue: please open an issue rather than editing a SKILL.md.

AI CTO core (MCP) · topy-ai-cto

SkillTierWhat it does
topy-cto-ci-watchcoreHow the AI CTO watches CI around the clock: a GitHub webhook per run, an email only on a failure or a recovery, with the failing jobs and the log lines.
topy-cto-connectcoreConnect the TOPY CTO MCP server and check what it can see: the GitHub App and its repositories, CI, alerts and service health.
topy-cto-goalscoreThe AI CTO's standing goals and its status report for the CEO: uptime, deploy frequency and change failure rate, each with source and date or 'no data'.
topy-cto-morning-checkcoreEach morning: service health, CI on the default branches, pull requests waiting and anything red — the one thing to fix first.

CI · topy-ai-cto-ci

SkillTierWhat it does
topy-cto-ci-failureconnectedExplain a failed GitHub Actions run from its jobs and logs: the step, the cause in the log's own words, flaky or real, and the fix — then re-run or open an issue.
topy-cto-ci-hygienestandaloneKeep CI fast and trustworthy: required checks, caching, a time budget, pinned actions, least-privilege tokens and no secrets in logs.
topy-cto-flaky-testsconnectedTell a flaky test from a real failure, quarantine it with an owner and a deadline, and fix the cause (time, order, network, shared state) instead of re-running forever.

Code review & issues · topy-ai-cto-code

SkillTierWhat it does
topy-cto-issue-triageconnectedTurn bug reports and CI findings into issues worth fixing: reproduce, size, label, owner — and close what isn't actionable.
topy-cto-pr-reviewconnectedReview a pull request for what matters before merge: what changed, whether checks pass, and the risks — migrations, auth, money, secrets, irreversible steps, missing tests.

Releases & deploys · topy-ai-cto-releases

SkillTierWhat it does
topy-cto-deploy-safelyconnectedShip to production in small steps: plan, the Founder's approval, deploy the gated build, canary, check live matches what was built, then report — or roll back.
topy-cto-release-gateconnectedDecide whether a build may ship: CI green on the exact commit, checks and QA passed, migrations reversible, rollback ready, and someone watching — evidence for each.
topy-cto-release-notesconnectedDraft release notes from the commits since the last release, grouped as new, improved and fixed in words customers understand — nothing is published.
topy-cto-rollbackconnectedWhen a release hurts, restore service first: redeploy the last good build, keep data safe, confirm recovery, then find the cause.

Dependencies & secrets · topy-ai-cto-security

SkillTierWhat it does
topy-cto-dependency-alertsconnectedRead Dependabot, code-scanning and secret-scanning alerts, fix what is exploitable first, and say plainly which alert kinds GitHub isn't showing and why.
topy-cto-secrets-hygienestandaloneKeep keys and tokens out of code, logs and chats: where secrets live, least privilege, rotation, and what to do in the first hour after one leaks.

Incidents & architecture · topy-ai-cto-platform

SkillTierWhat it does
topy-cto-architecture-decisionstandaloneRecord a technical decision so it can be revisited: the context, the options with their costs, the choice, its consequences, and what would make us change it.
topy-cto-incidentconnectedWhen something customers use is down or wrong: one lead, stop the harm, restore service, communicate on a clock, then a blameless review with fixes.

來源：README.md，提交 f0c0b29

工具

0
該伺服器需要登入，因此 SourceWeft 無法在沒有帳號的情況下讀取其工具，也未收錄。已連接帳號所能看到的內容可能因帳號及其權限而異。

版本歷史

1
  1. v0.1.0最新Oct 29, 2026

評分

已滿足適用檢查項的 85%

該評分衡量此收錄條目的完整程度，依據註冊中心條目、端點與公開來源計算得出。它不是對伺服器本身的評價，也不表示伺服器安全或目前可用：可連通性是帶日期的檢查結果，評分也不包含安裝量或使用者評價。

  • 近 30 天內可連通必需25 分
  • 已宣告認證方式必需10 分
  • 已列出工具15 分

    請讓伺服器在無憑證時回應 tools/list，或發布可在沙箱中無憑證啟動的軟體包。

  • 描述必需15 分
  • 圖示10 分
  • 文件15 分
  • 授權條款10 分

其他 Topy 整合

TOPY.AI · AI CEO

ai.topy官方發布者

TOPY.AI 的 AI CEO MCP 伺服器，讓助理撰寫並維護創辦人的商業計畫、CEO 紀錄、募資與財務工作。

Streamable HTTP可網頁執行驗證方式
效率與工作流需要憑證Oct 26, 2026

TOPY.AI · AI CSO

ai.topy官方發布者

TOPY.AI AI CSO 讓編碼助理具備資安長能力：CVE 與相依套件檢查、金鑰掃描、已驗證網站的網頁檢查、威脅模型與事件報告。

Streamable HTTP可網頁執行驗證方式
雲端與基礎設施需要憑證Nov 5, 2026

TOPY.AI · AI CRO

ai.topy官方發布者

遠端或本機執行的 MCP 伺服器，為助理提供小型公司的 AI CRO：依真實數據做營收檢視、銷售管道、交易台、報價與催收。

Streamable HTTP可網頁執行驗證方式
效率與工作流需要憑證Oct 29, 2026

TOPY.AI · AI CPO

ai.topy官方發布者

TOPY.AI AI CPO 讓助理具備產品管理能力：回饋分流、優先順序排序、PRD、預先註冊實驗與決策日誌。

Streamable HTTP可網頁執行驗證方式
效率與工作流需要憑證Oct 29, 2026

TOPY.AI · AI COO

ai.topy官方發布者

TOPY.AI 的 AI COO 伺服器，讓助理監控 AI 共同創辦人的排程 Routine、失敗、審批與每週營運報告。

Streamable HTTP可網頁執行驗證方式
效率與工作流需要憑證Oct 29, 2026