"Actual Budget" matching MCP servers:

Enables AI assistants to diagnose Crossplane control planes, discover resources and schemas, assess adoption and migration safety, and propose GitOps changes using evidence from actual cluster APIs.

AgentGuard47 is runtime safety infrastructure for AI agents. It adds budget caps, loop detection, retry limits, timeouts, local traces, and incident reports so agents can stop bad runs while they are happening, not just explain them afterward.

Governed AI-ops for managed-endpoint fleets, providing login-storm analysis and patch/config drift detection with built-in audit, budget, and risk-tier governance.

Tracks AI agent token usage and spending in real time, with budget alerts, per-task cost breakdown, and a visual dashboard.

Provides read-only, governance-audited operations for OLVM and oVirt 4.5 engines, enabling inventory, health, capacity, and diagnosis queries over MCP and CLI with audit logging and budget guards.

Governed, read-only OT data tap for substation and utility telecontrol — IEC 60870-5-104, DNP3/IEEE 1815, and IEC 61850 MMS connectors with cross-protocol asset discovery, alarm and downtime RCA, unbypassable audit logging (MCP + CLI), budget/runaway guards, and an airgap no-egress mode. Energy edition of Industrial-AIOps, built on iaiops.core.

Governed reverse-proxy operations for Traefik, Caddy, and HAProxy — backend-health, cert-expiry, 5xx, and route-conflict RCA, with unbypassable audit logging (MCP + CLI), budget/runaway guards, dry-run, and undo/rollback.

Governed MySQL + MariaDB DBA operations — slow-query, lock-wait/deadlock, replication, and fragmentation RCA, with unbypassable audit logging (MCP + CLI), budget/runaway guards, dry-run, and undo/rollback.

Policy sidecar for any MCP server: deny-by-default tool allowlist with argument rules, call and USD budget caps, and a hash-chained audit log with RFC 3161 timestamps.

Enables AI agents to ask what changed in a time window and get back a short, ranked answer across metrics, logs and traces, with the service where a failure started identified and its echoes in other services labelled as such. Read-only and deterministic, it scores changes against a baseline and trims results to a token budget, so agents avoid paging through raw queries and blaming the loudest symptom.

Governed Redis + RabbitMQ middleware operations — memory-pressure, latency, backlog, and churn RCA, with guarded queue/key writes, unbypassable audit logging (MCP + CLI), budget/runaway guards, dry-run, and undo/rollback.

Governed PostgreSQL DBA operations — slow-query, bloat, and blocking-lock RCA, index management, vacuum/analyze, and replication lag, with unbypassable audit logging (MCP + CLI), budget/runaway guards, dry-run, and undo/rollback.

Governed Docker + Portainer container-host operations (non-Kubernetes) — restart-loop, resource-pressure, and image/volume-bloat RCA, with guarded lifecycle writes, unbypassable audit logging (MCP + CLI), budget/runaway guards, dry-run, and undo/rollback.

Governed network controller-layer operations (Cisco Meraki, plus Catalyst Center / Arista CVP / UniFi) — uplink loss/latency RCA, fleet health scoring, and config-template drift, with unbypassable audit logging (MCP + CLI), budget/runaway guards, and rollback.

Governed Prometheus + Grafana operations — firing-alert and scrape-target RCA, alert noise/flapping analysis, silences, and dashboards, with unbypassable audit logging (MCP + CLI), budget/runaway guards, dry-run, and undo/rollback.

Governed MinIO object-storage operations — capacity, exposure-surface audit, lifecycle, and healing RCA, with guarded bucket writes, unbypassable audit logging (MCP + CLI), budget/runaway guards, dry-run, and undo/rollback.

The watchdog for unattended AI agents: flags MISSED, FAILED, NO_EVIDENCE, RETRY_STORM, BUDGET, DRIFT and STALLED runs of Claude Code routines, OpenClaw, n8n and cron jobs, and alerts via Telegram, Slack or webhook. MIT and self-hostable.

Enables developers to register sync or async MCP tools that are automatically protected by context-budget guards, including DataFrame blockades, payload truncation, and log-line capping. Each invocation is wrapped with correlated trace IDs and fail-closed error envelopes, and the server can be mounted as an SSE endpoint inside a FastAPI application.

MCP server for tracking LLM costs, setting spending limits, and receiving budget alerts directly within AI editors via slash commands like /guard.

Stop runaway agents. One gateway in front of all your agent traffic — MCP tools and LLM models — one budget, one kill switch. Sits between AI clients (Cursor, Claude Desktop, Windsurf) and remote MCP servers / OpenAI-compatible model APIs; hard token budgets; a single static Go binary.