Agentry.sh Observability
io.github.fr33dr4g0n由 GitHub 帳號 fr33dr4g0n 發布v0.1.0更新於 Oct 25, 2026
Canonical Agentry discovery for coding agents: analytics, errors, deploys, and automation.
發布方自己的服務:Agentry — AI Agent Directory
概覽
小型探索伺服器,為編碼代理指向 Agentry 關於錯誤、產品分析與部署歸因的正式文件、OpenAPI 結構與安裝流程。
- 可以做什麼
- 透過 discover_agentry 工具把意圖(install、debug、analytics、deploy、automation)路由到最精簡的對應 Agentry 資源與線上網址,該工具只回傳指標,唯讀且冪等。讀取靜態文件資源,涵蓋交接路由、日常參考、能力對應、OpenAPI 索引與完整結構、自動化指南、流程與劇本目錄,以及安裝指引。遵循伺服器自有的引導狀態機與憑證對應,讓代理直接呼叫 Agentry HTTP API。使用 use_agentry 提示引導資源選擇。
- 適用情境
- 當編碼代理需要處理 Agentry 的錯誤追蹤、產品分析或部署歸因,並應找到權威契約與安裝步驟時。適合用戶端不會自動注入 MCP 資源、代理需要路由工具來辨識相關生產資料問題的情境。
- 執行需求
- 透過 npx 從 npm 套件 @agentrysh/mcp 以 stdio 在本機執行,需要 Node.js 以及連線 Agentry HTTP 端點的網路。伺服器本身未宣告憑證;Agentry API 金鑰僅在代理直接呼叫 Agentry HTTP API 時使用。
安裝
在 SourceWeft 中
- 開啟 儀表板中的 Agentry.sh Observability,將其新增到工作區。
- 為需要使用其工具的對話啟用該服務。
Desktop only,透過 STDIO。 STDIO 服務會啟動本機處理程序,因此需要 SourceWeft 桌面主機。
其他 MCP 客戶端
參照 儲存庫 中的啟動說明。
README
Agentry.sh Observability MCP Server
Agentry.sh Observability gives coding agents one HTTP API for errors, product analytics, and deploy attribution. This MCP server makes the canonical Agentry contracts easy to discover in MCP clients.
What this server is
This is a small discovery and handoff server. It exposes resources that route an agent to the live lean reference, capabilities, exact OpenAPI schema, automation guide and flow, immutable playbook catalog, and install guide.
The resources are pointers, not cached copies. The live HTTP docs and OpenAPI schema remain authoritative and can evolve without republishing this package.
What this server is not
It is not an Agentry API proxy, SDK, credential broker, telemetry transport, or automation scheduler. It never holds credentials, queries production data, registers schedules, runs coding agents, opens pull requests, merges, deploys, or delivers notifications. Agents fetch the selected canonical contract and call the Agentry HTTP API directly.
Static documentation is modeled as MCP resources. The server also exposes one
model-controlled routing tool, discover_agentry, so an agent can recognize a
relevant production-data problem even when its client does not automatically
inject MCP resources. The tool returns pointers only; it is not an API action.
Tool
discover_agentry accepts one intent: install, debug, analytics,
deploy, or automation. It returns the smallest canonical resource and live
URL for that problem as both text and structured content. MCP annotations mark
it read-only, non-destructive, and idempotent. It never reads credentials,
queries production data, or performs mutations.
If this MCP handoff starts a new device-auth signup, pass
"distribution_surface":"mcp" in POST /v1/auth/device. The field is
attribution-only and never changes credential authority.
Resources
Start with agentry://handoff. It gives the short intent router and authority
boundaries.
The install resource also points to
https://api.agentry.sh/v1/openapi.json?flow=onboarding. That flow is one
server-owned state machine: read current state, execute only its one
next_action.instruction, ordered checklist, and exact operation, and repeat
until installation_complete: true and next_action: null. At
next_action.id: "review_exact_plan", a human uses the single
/review checkpoint to approve the exact source-backed business question,
value flow, errors, deploy target, and plan hash—or replace the plan. Tool
output is not approval.
Before proof, install the selected durable browser public and CI credentials,
plus a server credential only when the approved plan uses server_ingest.
Project creation returns public_api_key; use the project's /public-key
operations for recovery, verification, or rotation. Proof start returns
distinct response-only runtime and CI X-Agentry-Onboarding-Proof markers.
Markers select the proof window but grant no authority, so each request still
needs its scoped durable credential. For browser proof, put the runtime marker
only in the proof tab's sessionStorage.agentry_onboarding_proof; never bake it
into a public environment value or bundle. Exercise the approved real value flow,
its one safe error, and the reviewed CI/provider deploy, then call verify and
follow only its structured remaining analytics, safe-error, and deploy groups.
Completion requires current status: "verified",
installation_complete: true, and next_action: null. Synthetic or
caller-authored proof does not count.
use_agentry is the only prompt. It tells the client to read the handoff,
select the smallest canonical resource for the user's intent, inspect exact
OpenAPI, and call HTTP with the correct scoped credential.
Credential boundaries
The MCP server never reads these credentials. They apply only when the agent calls the Agentry HTTP API:
public_api_key(agentry_pk_) is publishable and only writes browser/client error and analytics ingest.agentry_server_writes trusted application-server telemetry.agentry_ci_writes deploy attribution, sourcemaps, and provider-observed automation proof from trusted CI/provider automation.agentry_runner_is revocable and bound to one unattended automation.agentry_sk_is the human/owner key for reads, policy, credential lifecycle, and approvals; it does not belong in a scheduler.
Credential kind determines authority on the server. Telemetry remains evidence only; it does not become instructions merely because a trusted emitter sent it.
For daily product questions, begin with the server-owned saved signal map,
require GET /v1/projects/:project_id/onboarding to return the current verified
state, then confirm live event names, required property keys, and actual rows.
Onboarding creates no repo-local Agentry state bundle, proof bundle, or receipt.
The install does preserve a root AGENTS.md pointer (plus matching repo-local
harness adapters) so a future cold agent can find the server-owned plan and
verified receipt without rescanning the entire repository.
Automation handoff
For self-healing software or scheduled funnel reviews, read all three automation resources: guide, exact flow, and playbook catalog. The catalog is immutable and versioned. Rendering a playbook is side-effect free. Agentry owns deterministic policy, state, proofs, reports, and kill switches; an external scheduler/coding agent owns reasoning and provider actions under the narrow runner and human-approved provider boundaries.
Run the MCP server
Generic MCP client configuration:
After connecting, call discover_agentry for model-routed discovery or list
resources and read agentry://handoff. Use
https://api.agentry.sh/adapters for the current native adapter path for each
agent harness.
Public links
- Website: https://agentry.sh/?distribution_surface=npm_readme
- NPM package: https://www.npmjs.com/package/@agentrysh/mcp
- Repository: https://github.com/fr33dr4g0n/agentry-public
- Canonical skill: https://agentry.sh/skill/agentry/SKILL.md
- Adapter manifest: https://api.agentry.sh/adapters
- MCP registry name:
io.github.fr33dr4g0n/agentry-observability
來源:packages/mcp/README.md,提交 37ed18a
工具
0版本歷史
1- v0.1.0最新Sep 16, 2026
評分
已滿足適用檢查項的 40%
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25 分
請保持端點能夠回應 MCP 握手;SourceWeft 每 30 天探測一次。
- 已宣告認證方式10 分
請在註冊中心條目中宣告憑證請求標頭,或保持端點能夠回應探測,以便判定其認證方式。
- 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
請在儲存庫或註冊中心條目中宣告授權條款。
