BIGHUB

io.bighub由 bighub.io 發布v0.2.1更新於 Oct 24, 2026

Decision learning for AI agent actions. Evaluate, score, decide, and learn from outcomes.

已驗證STDIO僅桌面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

BIGHUB 的決策層：助理可提交擬執行的 IT 動作，取得執行、複核或阻擋的決策，並回報結果。

可以做什麼
  • 評估擬執行的 IT 動作，回傳含執行模式、can_run、複核與風險標記的正規化決策
  • 建立決策封包、執行 DecisionBrain，並處理待人工複核的請求
  • 回報結果，查詢決策案例、先例、校準與檢索訊號
  • 檢視洞察、模擬、特徵、學習執行與執行期事件擷取
  • 管理 Okta、Slack、GitHub、雲端與 CI 等系統整合及其輪詢
  • 管理營運限制、核准、終止開關、Webhook、API 金鑰與登入驗證
適用情境
適用於助理提出具實際後果的 IT 動作時，例如授予存取權、輪換憑證、部署或回復服務、變更 IAM 或 CI/CD 設定、匯出敏感資料，並希望在執行前取得決策與複核關卡。僅回答問題或只做唯讀查詢的助理不需要它。
執行需求
以 Node.js 18+ 透過 npx 在本機以 stdio 方式執行。需要 BIGHUB 帳號，以及 BIGHUB_API_KEY 或 BIGHUB_BEARER_TOKEN 其中之一；API 金鑰、Webhook 與驗證等管理工具需要 bearer token。選用設定包括 BIGHUB_BASE_URL、逾時、重試與允許不安全 HTTP 的開關。需要能連線至 BIGHUB API 的網路。
安裝前請注意
伺服器持有 BIGHUB API 金鑰或 bearer token，並將擬執行動作與情境傳送至 BIGHUB API。許多工具會寫入或改變狀態：處理複核與核准、建立或刪除限制、連線、Webhook 與 API 金鑰、輪換金鑰、啟用終止開關、觸發學習或輪詢。另提供通用 HTTP 透傳工具，可存取任意端點。啟用前請確認助理可執行的範圍。

安裝

在 SourceWeft 中

  1. 開啟 儀表板中的 BIGHUB，將其新增到工作區。
  2. 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

@bighub/bighub-mcp

BIGHUB's Better Decision layer as an MCP server. Turn proposed IT agent actions into better decisions before they run from any MCP-compatible client.

MCP server for better decisions on IT agent actions.

text
MCP client   ↓@bighub/bighub-mcp   ↓BIGHUB API   ↓proposed action → Decision Packet → DecisionBrain → decision signals → run / review / context / block

Table of contents

Start here

Tool reference

Reference

Install

bash
npm install @bighub/bighub-mcp

Requires Node.js 18+.

Quickstart

Set your API key:

bash
export BIGHUB_API_KEY=your_api_key

Run the server in stdio mode:

bash
npx @bighub/bighub-mcp

Add it to your MCP client configuration:

json
{  "mcpServers": {    "bighub": {      "command": "npx",      "args": ["@bighub/bighub-mcp"],      "env": {        "BIGHUB_API_KEY": "your_api_key"      }    }  }}

Works with MCP-compatible clients such as Claude Desktop and Cursor.

When to use bighub-mcp

Use this server when your MCP-connected agent proposes IT actions such as:

  • granting Okta access
  • rotating credentials
  • deploying or rolling back services
  • changing CI/CD, cloud, or IAM settings
  • exporting sensitive data
  • posting incident or support updates
  • resolving actions that need human review

If your agent only answers questions or performs read-only lookups, you don't need this. It's designed for agents that make decisions with real-world consequences.

Typical MCP Loop

  1. Send the proposed IT action to bighub_decide
  2. Let BIGHUB build or use a Decision Packet
  3. Run DecisionBrain and return execution mode, review/context flags, and a decision path when available
  4. Execute only when can_run is true and no blocking flags are set
  5. Route to review when BIGHUB asks for human approval or modification
  6. Optionally report the outcome later
text
bighub_decide→ bighub_build_packet / bighub_run_brain→ bighub_resolve_review when needed→ bighub_report_outcome later, optional

Better Decision primitives

The modern MCP surface is centered on the decision flow:

FieldDescription
bighub_decideTurn a proposed IT action into a better decision
bighub_build_packetBuild a Decision Packet from action and context
bighub_run_brainRun DecisionBrain on a packet
bighub_list_reviewsList pending reviews
bighub_resolve_reviewApprove, deny, or modify the reviewed action
bighub_get_system_contextFetch context for Okta, Slack, or a configured integration provider
bighub_get_world_stateRead operational world state
bighub_report_outcomeOptionally report what happened later
bighub_http_requestLow-level fallback for endpoints not yet modeled

bighub_decide and bighub_run_brain return a normalized decision object first: recommended_action, mode, can_run, needs_review, needs_more_context, should_not_run, risk, confidence, expected_regret, reason, system, selected_model when available, decision_path when available, and counts for verification steps and obligations. The original backend payload is still available under raw for advanced callers. Legacy tools such as bighub_actions_evaluate remain available for compatibility.

bighub_build_packet uses the same canonical JSON SHA-256 hashing strategy as the Python SDK, so packet hashes match across SDK and MCP clients.

Trajectory-aware evaluation

BIGHUB evaluates actions not only in isolation, but also in the context of what happened before. As deployment, review, and outcome history accumulates, similar IT actions can receive better recommendations over time.

Complete Tool Reference

BIGHUB tools are organized by domain. The core loop tools are listed first.

Core loop

ToolDescription
bighub_decideModern Better Decision entry point
bighub_build_packetBuild a Decision Packet
bighub_run_brainRun DecisionBrain
bighub_report_outcomeOptionally report what happened later
bighub_actions_evaluateLegacy / low-level action evaluation

Actions

ToolDescription
bighub_actions_evaluateLegacy / low-level action evaluation
bighub_actions_evaluate_payloadEvaluate with a free-form payload
bighub_actions_evaluate_batchEvaluate multiple actions in one request
bighub_actions_dry_runNon-persistent evaluation (preview)
bighub_actions_live_connectOpen a live connection slot
bighub_actions_live_heartbeatHeartbeat a live connection
bighub_actions_live_disconnectClose a live connection
bighub_actions_verify_validationVerify a validation hash
bighub_actions_observer_statsObserver statistics
bighub_actions_dashboard_summaryDashboard summary metrics
bighub_actions_statusService status
bighub_actions_memory_ingestIngest decision memory events
bighub_actions_memory_contextRetrieve memory context
bighub_actions_memory_refresh_aggregatesRefresh memory aggregates
bighub_actions_memory_recommendationsPattern-based recommendations from memory

Outcomes

ToolDescription
bighub_outcomes_reportReport a real-world outcome
bighub_outcomes_report_batchBatch report outcomes
bighub_outcomes_getGet outcome by request_id
bighub_outcomes_get_by_validationGet outcome by validation_id
bighub_outcomes_get_by_caseGet outcome by case_id
bighub_outcomes_timelineFull outcome timeline
bighub_outcomes_pendingDecisions awaiting outcomes
bighub_outcomes_analyticsOutcome analytics summary
bighub_outcomes_taxonomySupported outcome statuses
bighub_outcomes_recommendation_qualityFollow rate, quadrants, trend, by domain/actor
bighub_outcomes_partner_viewSelf-contained domain view with KPIs and examples

Decision cases

ToolDescription
bighub_cases_createCreate a decision case
bighub_cases_getGet a case by ID
bighub_cases_listList and filter cases
bighub_cases_report_outcomeReport outcome for a case
bighub_cases_precedentsPrecedent intelligence for a proposed action
bighub_cases_calibrationCalibration metrics for cases

Precedents

ToolDescription
bighub_precedents_queryQuery similar past cases
bighub_precedents_signalsAggregated precedent signals
bighub_precedents_statsPrecedent index statistics

Calibration

ToolDescription
bighub_calibration_reportCalibration report
bighub_calibration_reliabilityReliability diagram data
bighub_calibration_driftCalibration drift over time
bighub_calibration_breakdownBreakdown by domain or tool
bighub_calibration_feedbackCalibration feedback signals
bighub_calibration_observeSubmit a calibration observation
bighub_calibration_quality_historyDaily quality score over time

Multi-signal retrieval

ToolDescription
bighub_retrieval_queryMulti-signal precedent retrieval
bighub_retrieval_query_explainedRetrieval with explanation trace
bighub_retrieval_strategiesList available strategies
bighub_retrieval_strategyGet one strategy's details
bighub_retrieval_index_caseManually index a case
bighub_retrieval_compareCompare two strategies
bighub_retrieval_statsRetrieval index statistics

Insights

ToolDescription
bighub_insights_adviseLearned advisories for an action
bighub_insights_patternsDiscovered risk patterns
bighub_insights_learnLearning refresh
bighub_insights_profileAction/tool profile

Simulations

ToolDescription
bighub_simulations_listList simulation snapshots
bighub_simulations_getGet a snapshot
bighub_simulations_by_requestGet snapshot by request
bighub_simulations_compareCompare predicted vs actual
bighub_simulations_accuracyDomain-level accuracy
bighub_simulations_statsSimulation statistics

Learning

ToolDescription
bighub_learning_strategyCurrent learning strategy
bighub_learning_runsRecent learning runs
bighub_learning_recomputeTrigger learning recomputation
bighub_learning_backfillBackfill learning artifacts

Features

ToolDescription
bighub_features_computeCompute features for an action
bighub_features_compute_batchBatch feature computation
bighub_features_snapshotCreate a feature snapshot
bighub_features_get_snapshotGet a snapshot
bighub_features_list_snapshotsList snapshots
bighub_features_explainExplain features for a case
bighub_features_exportExport features for a case
bighub_features_export_batchBatch export
bighub_features_compareCompare feature sets
bighub_features_schemaFeature schema
bighub_features_statsFeature statistics

Runtime ingestion

ToolDescription
bighub_ingest_eventIngest a single event
bighub_ingest_batchIngest events in batch
bighub_ingest_reconcileReconcile outcome with event
bighub_ingest_lifecyclesList event lifecycles
bighub_ingest_lifecycleGet lifecycle for one event
bighub_ingest_pendingEvents pending reconciliation
bighub_ingest_staleStale unreconciled events
bighub_ingest_statsIngestion statistics
bighub_ingest_adaptersList available adapters

System integrations

ToolDescription
bighub_systems_list_connectionsList configured system integrations
bighub_systems_get_connectionGet one provider connection summary
bighub_systems_test_connectionTest unsaved provider config
bighub_systems_save_connectionSave provider config with optional display_name
bighub_systems_delete_connectionDelete provider connection
bighub_systems_poll_nowTrigger one provider poll now
bighub_systems_get_poll_scheduleGet provider poll schedule
bighub_systems_update_poll_scheduleConfigure provider polling
bighub_systems_poll_historyRead redacted provider poll history
bighub_systems_poll_statusScheduler status and due counts
bighub_systems_poll_metricsProvider success/failure, latency, stale schedule and verifier metrics
bighub_systems_run_due_pollsRun due polls for the current organization

Supported system providers include okta, slack, github, sentry, datadog, aws_cloudtrail, terraform, kubernetes, argocd, gitlab, jenkins, azure, prometheus, grafana, and openshift. Common aliases such as github-ci, cloudtrail, terraform-cloud, k8s, argo-cd, gitlab-ci, and ocp are normalized to backend provider IDs.

Poll snapshots and histories are redacted before persistence and exposure. They provide best-effort operational context; they are not a replacement for production monitoring.

Operating constraints

ToolDescription
bighub_constraints_createCreate a constraint
bighub_constraints_listList constraints
bighub_constraints_getGet a constraint
bighub_constraints_updateUpdate a constraint
bighub_constraints_deleteDelete a constraint
bighub_constraints_pausePause a constraint
bighub_constraints_resumeResume a constraint
bighub_constraints_dry_runPreview without persisting
bighub_constraints_validateValidate an action against constraints
bighub_constraints_validate_dry_runValidate (dry run)
bighub_constraints_domainsList domains with constraints
bighub_constraints_versionsConstraint version history
bighub_constraints_apply_patchApply a JSON Patch
bighub_constraints_purge_idempotencyAdmin: purge idempotency keys

Approvals & kill switch

ToolDescription
bighub_approvals_listList approval requests
bighub_approvals_resolveResolve an approval
bighub_kill_switch_statusKill switch status
bighub_kill_switch_activateActivate kill switch
bighub_kill_switch_deactivateDeactivate kill switch

Events

ToolDescription
bighub_events_listQuery event stream
bighub_events_statsEvent statistics

Webhooks

ToolDescription
bighub_webhooks_createCreate a webhook
bighub_webhooks_listList webhooks
bighub_webhooks_getGet a webhook
bighub_webhooks_updateUpdate a webhook
bighub_webhooks_deleteDelete a webhook
bighub_webhooks_deliveriesList deliveries
bighub_webhooks_testSend a test delivery
bighub_webhooks_list_eventsList subscribable event types
bighub_webhooks_verify_signatureVerify webhook signature
bighub_webhooks_replay_failed_deliveryReplay a failed delivery

API keys

ToolDescription
bighub_api_keys_createCreate an API key
bighub_api_keys_listList API keys
bighub_api_keys_deleteDelete an API key
bighub_api_keys_rotateRotate an API key
bighub_api_keys_validateValidate an API key
bighub_api_keys_scopesList available scopes

Auth

ToolDescription
bighub_auth_signupCreate an account
bighub_auth_loginLog in
bighub_auth_refreshRefresh token
bighub_auth_logoutLog out

Utility

ToolDescription
bighub_http_requestGeneric HTTP passthrough

Environment Variables

VariableRequiredDefaultDescription
BIGHUB_API_KEYYes*-API key auth
BIGHUB_BEARER_TOKENNo-Alternative bearer auth
BIGHUB_BASE_URLNohttps://api.bighub.ioAPI base URL
BIGHUB_TIMEOUT_MSNo15000Request timeout
BIGHUB_MAX_RETRIESNo2Retries on transient failures
BIGHUB_ALLOW_INSECURE_HTTPNo-Allow HTTP for local/private testing

* One of BIGHUB_API_KEY or BIGHUB_BEARER_TOKEN is required.

Management tools (API keys, webhooks, auth) require user JWT auth via BIGHUB_BEARER_TOKEN.

Local Development

bash
git clone https://github.com/bighub-io/bighub.gitcd bighub/servers/mcpnpm installnpm run testnpm run checknpm run buildnpm run startnpm run dev

Links

License

Apache-2.0

來源：servers/mcp/README.md，提交 61f4e2f

工具

0
SourceWeft 尚未在沙箱中啟動過該軟體包，因此其工具未知。

版本歷史

1
  1. v0.2.1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。