Skip to main content
Kubernetes AI Operations Framework|Open Source

The AI Operations Framework for Kubernetes.

k8s-autopilot is an open-source, extensible AI operations framework built on LangGraph and the Deep Agents SDK. Talk to your cluster in natural language to manage Helm releases, orchestrate progressive Argo Rollouts, troubleshoot crashing pods, and correlate full-stack observability — all protected by Human-in-the-Loop governance and autonomous rubric verification.

k8s-autopilot
Autonomous Agent
k8s-autopilot — Open-Source AI Operations Framework for Kubernetes
One-line Install
$curl -LsSf https://raw.githubusercontent.com/talkops-ai/k8s-autopilot/main/scripts/install.sh | bash
Apache 2.0LangGraph 1.2A2A ProtocolPython 3.12+Docker Ready
4 Operators
Built-in Domains
Helm, App, K8s, Obs
11 Built-in
MCP Server Support
Zero-code tool binding
3 Modes
HITL Governance
Manual, Auto, & YOLO
20+ Models
Universal Providers
Claude, OpenAI, Gemini
Plugins
Extensibility Factory
Claude, Codex, OpsCode
A2UI Native
Interactive Surfaces
Visual cards & streaming
Core Capabilities

Extensible Architecture for Cloud-Native Operations

An open AI operations framework built on LangGraph. From out-of-the-box Kubernetes operators to custom sub-agents and plugins, automate complex infrastructure workflows with strict safety and rubric verification.

Multi-Agent Framework

Dynamic Multi-Agent Supervisor Architecture

An open, extensible AI operations runtime built on LangGraph and the Deep Agents SDK — not a closed tool. A central Supervisor coordinates foundational out-of-the-box operators alongside dynamically spawned custom sub-agents, binding MCP tool servers Just-In-Time (JIT) to scale across any domain without token bloat.

Supervisor-to-Operator delegation with persistent state handoffs
Out-of-the-box K8s operators, infinitely extendable to any cloud or stack
Just-In-Time (JIT) tool binding: MCP servers spin up strictly on-demand
Human in the Loop

Predictable Governance & 3 Safety Modes

Look before you touch. Control operational autonomy across 3 runtime modes: Manual (every mutating action gated), Auto (3-tier safety filter: pattern checks → AI risk evaluation → human gate), and YOLO. Always fails safe, with dry-run previews and elevated scrutiny on production workloads.

3 Runtime modes: Manual, Auto, & YOLO switchable from UI
3-Tier safety filter with automatic human approval fallback
Pre-execution dry-run manifests & blast-radius inspection
Autonomous Verification

Goal Tracking & Rubric Grader Loops

Propose high-level cluster outcomes in natural language. The agent auto-generates concrete, verifiable criteria, then executes in an iterative Worker + Grader loop — an independent model evaluates actual cluster state until 100% of criteria pass before declaring completion.

Dual-model Worker + Grader loop with verifiable pass/fail rubrics
Auto-generates multi-step acceptance checklists from natural language
Autonomous CI/CD and self-healing remediation validation
Universal Extensibility

Plugin Marketplace & Custom MCP Servers

Extend the runtime to any tool or cloud with zero code modifications. Discover and install plugins from remote marketplaces, GitHub, or local catalogs to add skills, new sub-agents, and tools. Natively supports Claude, OpenAI Codex, and OpsCode plugin formats with hot-reload MCP integration.

Marketplace: Install skills, sub-agents, & MCP tool connectors
Universal compatibility with Claude, OpenAI Codex, & OpsCode plugins
Hot-reload custom MCP servers from the UI with zero restarts
Interactive Surfaces

A2UI & A2A Native Visual Interfaces

Say goodbye to walls of raw terminal text. Powered by the open Agent-to-User (A2UI) and Agent-to-Agent (A2A) protocols, workflows stream into rich visual surfaces: interactive one-click approval cards with blast-radius diffs, live step execution trackers, and real-time metric charts.

Interactive visual approval cards with risk badges & diff previews
Real-time tool execution progress with streaming output drawers
Standardized A2A protocol for cross-agent & multi-cluster orchestration
Why k8s-autopilot

Engineered Specifically for Cloud-Native Kubernetes Ops

Traditional AI assistants either lack live cluster understanding or force vendor lock-in. k8s-autopilot combines multi-operator reasoning, deep tool isolation, and rigorous safety verification.

CapabilityKubiyaKomodorBotkubekagent
k8s-autopilotOPEN FRAMEWORK
Open Source LicenseProprietaryProprietaryPartialCNCF Sandbox
Apache 2.0 (100% Open Source)
Multi-Agent ArchitectureYes (Proprietary)No (Single)No (Bot-based)Yes (A2A per-CRD)
Supervisor + Pluggable Operators & Sub-agents
MCP Tool Server EcosystemNoneNoneNoneGeneric MCP
11 Built-in Servers + Dynamic Custom MCP
Plugin Marketplace FormatsLimitedNoneInternal CatalogBYO Tool Servers
Claude, Codex, & OpsCode Formats
Built-in Domain CoverageWorkflow ActionsK8s TroubleshootEvent AlertsGeneral Runtime
Helm, Argo, Rollouts, Traefik, Prom, Loki, Tempo
Human-in-the-Loop GovernanceJIT ApprovalsApproval FlowsChat CommandsCRD Gates
Manual, Auto (AI Classifier), YOLO + A2UI
Goal Tracking & Rubric GradingNoNoNoNo
Worker + Grader Self-Evaluating Loops
Universal LLM ProvidersUndisclosedUndisclosedOpenAI Only5 Providers
20+ Providers with Live UI Hot-Swap
Architecture & Extensibility

Deep Multi-Operator Architecture. Governed by Design.

Built on LangGraph and the Deep Agents SDK. Explore the architectural foundations that make k8s-autopilot the most extensible, safe, and transparent AI operations framework for Kubernetes.

Multi-Agent Orchestration

Supervisor & 4 Domain Operators

Rather than forcing a single model to do everything, k8s-autopilot uses a central Supervisor that classifies intent and delegates tasks to specialized Domain Operators. Each operator runs in an isolated context loop with dedicated MCP servers.

  • Clean Context Sandboxing: Intermediate command outputs, dry-runs, and PromQL query iterations stay contained inside the operator — never polluting your conversation.
  • Just-In-Time (JIT) Tooling: MCP tool servers start as lightweight subprocesses only when needed and terminate immediately after task completion.
  • Cross-Operator Structured Handoffs: When investigations cross domains (e.g. Observability detects latency → K8s Operator inspects crashing pods), full context is preserved.
Operator Inspector & MCP Bindings
MULTI-AGENT
Helm Operator
Helm Chart Lifecycle & Schema Validation
helm-mcp-server (22 tools)
helm-operationvalues-schema-checkdry-run-validationrevision-rollback
// Helm Operator (helm-mcp-server · 22 tools)
Task: Install Redis cluster from bitnami in namespace [database]
Pipeline: Discover → Plan → Confirm → Execute → Verify
Safety: Values schema validated | helm install --dry-run verified
Approval Gate: Pinned rollback checkpoint to exact revision #3
Safety & Governance

Human-in-the-Loop: Look Before You Touch

k8s-autopilot follows a strict co-pilot rule: read-only actions run immediately, while any action that could change infrastructure pauses and presents an interactive approval card before executing.

DEFAULT FOR PRODUCTION

Every state-modifying action (deploy, upgrade, delete, scale, or traffic route) pauses and presents an interactive Approval Card with target resources and blast radius. Nothing touches your cluster without explicit confirmation.

Security Pipeline: Manual Mode
4-TIER GUARD
01
Shell Pattern & AST Parser
PIPES & CHAINS AUDITED
02
Mutating Action Gate
INTERCEPTED & PAUSED
03
A2UI Approval Card
HUMAN CONFIRMATION
04
Cluster Execution
DISCOVER → PLAN → EXECUTE
Safety Pattern: Discover → Plan → Confirm → Execute → Verify
Verification & Quality

Goals & Rubrics: Verifiable Execution

When given multi-step infrastructure tasks, k8s-autopilot doesn't execute blindly. It proposes concrete acceptance criteria that become an active rubric, running a Worker + Grader loop to verify cluster state.

  • Objective Verification: Task completion is evaluated by an independent rubric grader model against actual cluster resources rather than self-assessed.
  • Iterative Self-Healing: Failed criteria generate a structured deficiency report, and the worker automatically remedies gaps until all criteria pass.
  • Tactical Todo Checklist: Execution tracks in real-time through pending, in_progress, and completed states, keeping the agent anchored during long tasks.
Attempt #1
Worker generates Redis StatefulSet and ClusterIP Service
Grader Score: 2/4 (Deficiency Report Generated)
Missing: ServiceMonitor CRD for Prometheus & StorageClass PVC binding
↓ Auto-Remediation with Grader Feedback ↓
Attempt #2
Worker adds redis-exporter sidecar, ServiceMonitor, and PVC volumeClaimTemplates
Grader Score: 4/4 (ALL CRITERIA VERIFIED)
All 4 acceptance criteria verified on live cluster. Goal completed!
Extensibility & Plugins

Agent & Vertical Plugins with Universal Formats

Extend k8s-autopilot without modifying core agent code. Install add-ons with one click from the Settings UI to add dedicated sub-agents, workflow playbooks, or custom MCP tool connections.

  • Two Plugin Types: Agent Plugins register dedicated sub-agents in isolated scopes; Vertical Plugins add skills and commands directly to the main agent.
  • Universal Marketplace Formats: Works natively with plugins built for Anthropic (Claude), OpenAI (Codex), OpsCode, or native k8s-autopilot format.
  • Live Settings UI at :8888: Model selection, credentials, cluster contexts, and MCP servers configure at runtime with zero restarts.
Configuration Precedence Cascade
HIERARCHY
1. HIGHESTDatabase (Settings UI at :8888)
Runtime configurable, persistent SQLite/PostgreSQL, zero restarts
2. GLOBALGlobal User Config (~/.k8s-autopilot/)
Developer workstation defaults across clusters and sessions
3. PROJECTProject Config (.k8s-autopilot/)
Git-versioned repository settings shared across engineering teams
4. PLUGINSMarketplace Plugins (Claude, Codex, OpsCode)
Agent and vertical plugins from GitHub, URLs, or local paths
5. DEFAULTBuilt-In Operators & 11 MCP Servers
Helm, ArgoCD, Rollouts, Traefik, K8s, Prometheus, Loki, Tempo
Model Independence

20+ Model Providers. Zero Lock-In.

Switch model providers on the fly directly from the Settings UI. API keys are saved to your configuration database with zero server restarts required.

Frontier Reasoning & Foundation

Direct API & Cloud
Anthropic
Claude Opus 5 / 4.8 / 4.7, Sonnet 5 / 4.6, 3.7 Sonnet (Thinking up to "max" effort · 16K budget)
OpenAI
GPT-5.5 Pro, GPT-5.5, GPT-5.4 Pro, o3 (Up to "xhigh" effort · 400K ctx · 128K output)
Google GenAI
Gemini 3.8 / 3.7 Flash (Thinking), 3.1 Pro (1M ctx · 64K output · Native thoughts)
DeepSeek
DeepSeek R1 (deepseek-reasoner), DeepSeek V4 Pro, DeepSeek V3 (Native CoT)
xAI
Grok 4.5, Grok 2 (131K context · Native reasoning effort control)

Enterprise Cloud & Sovereignty

VPC & IAM Secured
Google Vertex AI
Gemini 3.8 / 3.7 Flash, Gemini 2.5 Pro, Claude 3.7 Sonnet (ADC & Workload Identity)
AWS Bedrock
Claude Opus 4.7, Claude 3.7 Sonnet, Amazon Nova Pro, Llama 3.3 (Native IAM Roles)
Azure OpenAI
GPT-5.4, GPT-4o, o3-mini (Private VNet, Microsoft Entra ID & Managed Identity)
IBM watsonx
Granite 3 8B, Llama 3.3 70B (Enterprise data governance & compliance boundaries)
NVIDIA NIM
Nemotron 3 Ultra 550B, Llama 3.3 70B (Self-hosted microservices & GPU clusters)

High-Throughput, Gateways & Local

Zero Egress & Air-Gapped
Ollama
DeepSeek R1 (14B), Qwen 2.5 Coder, Llama 3.3 / 3.1 (100% offline, zero egress, sovereign)
Groq
Llama 3.3 70B, DeepSeek R1 Distill 70B (Deterministic LPU ultra-low latency streaming)
OpenRouter
Claude Opus 4.8/4.7, Claude Sonnet 5, GPT-5.5, DeepSeek V4 Pro, Kimi K3 (1M ctx), GLM 5.2
Fireworks & Together
GLM 5.2 (effort to max), DeepSeek V4 Pro, Kimi K3, Llama 3.3 Turbo, Qwen 2.5 72B
Mistral AI & LiteLLM
Mistral Large, Codestral (256K ctx), Pixtral Large & unified enterprise LiteLLM proxy
Dynamic Multi-Model Routing Strategy: Route high-frequency telemetry, log triage, and routine queries to fast inference tiers (such as gemini-3.7-flash, gpt-5.4-mini, or llama-3.3-70b), and escalate complex multi-repo GitOps rollouts, autonomous incident postmortems, and rubric evaluations to frontier reasoning engines (such as claude-opus-4.7/5 with max thinking effort, gpt-5.5-pro with xhigh reasoning, or deepseek-r1).
Real-World Automation

Governed Kubernetes Automation in Action

From zero-downtime progressive rollouts and full-stack observability triage to schema-validated Helm upgrades and Day-2 cluster governance, k8s-autopilot executes battle-tested infrastructure playbooks.

Observability + K8s OperatorsPrometheus · Tempo · Loki · K8s
CROSS-PILLAR INCIDENT TRIAGE

Production Incident Forensics & Auto-Remediation

Detects and resolves cluster incidents across all 5 telemetry pillars. Correlates Prometheus 5xx rate spikes and latency degradation with Tempo distributed trace spans, isolates matching error lines in Loki, diagnoses pod CrashLoopBackOff or OOMKilled events, and prepares non-destructive remediation manifests with rollback safety.

EXECUTION PLAYBOOK PIPELINE:
PromQL AnomalyTempo TraceQLLoki LogsPod DiagnosticsHITL Patch
Read-Only Discovery · Gated Cluster Mutations
PromQL/TraceQL Incident Brief + Memory Patch
App OperatorArgoCD · Argo Rollouts · Traefik
PROGRESSIVE DELIVERY & GITOPS

ArgoCD GitOps Sync & Canary Traffic Steering

Automates end-to-end application delivery pipelines. Converts standard Deployments to Argo Rollouts using workloadRef without duplicating pods or causing GitOps drift, configures Traefik IngressRoute for weighted traffic splits (e.g. 95/5 → 50/50), validates telemetry via AnalysisTemplate, and requires human sign-off before 100% promotion.

EXECUTION PLAYBOOK PIPELINE:
Repo OnboardingworkloadRef MigrationTraefik 80/20 RoutingAnalysisTemplate50%+ Human Gate
Autonomous Gating at 50% · Auto-Rollback on Anomaly
Rollout CRD + Traefik IngressRoute + ignoreDiffs
Helm Operator22 Tools · helm-operation Playbook
PACKAGE LIFECYCLE & HYGIENE

Helm Lifecycle & Schema-Validated Upgrades

Discovers and manages charts from public or private OCI registries. Parses values.schema.json to identify required fields, auto-detects existing releases to prevent naming collisions, preserves custom configurations via --reuse-values, and previews non-destructive dry-run manifest diffs before applying atomic updates.

EXECUTION PLAYBOOK PIPELINE:
Registry Discoveryvalues.schema.jsonUpgrade Auto-DetectDry-Run DiffRevision Rollback
Strict Schema Validation · Exact Revision Rollback History
Chart.yaml + values.yaml Diff + Revision Log
K8s Operatorkubernetes-cluster-ops Playbook
FLEET GOVERNANCE & DAY-2 OPS

Multi-Cluster Day-2 Ops, RBAC Audits & Scaling

Executes safe day-2 cluster operations across staging and production fleets. Audits RBAC bindings and service account privileges, diagnoses node CPU/memory pressure (nodes_top), monitors cluster-wide WarningEvents, safely scales workloads during load spikes, and launches isolated temporary debug pods—with automated secret masking.

EXECUTION PLAYBOOK PIPELINE:
Fleet Context SwitchWarningEvents TriageRBAC Privilege AuditNode Capacity CheckGoverned HPA Scale
Secret Masking · Prod & kube-system Safety Guardrails
RBAC Audit Matrix + Node Diagnostic Report + HPA Spec
Get Started in 60 Seconds

Ready to Bring Governed AI to Your Kubernetes Clusters?

Deploy the lightweight k8s-autopilot Docker Compose stack on Linux, macOS, or Windows (WSL). Zero manual configuration required to launch the interactive web dashboard and connect your clusters.

One-Command Installation (curl | bash)
$ curl -LsSf https://raw.githubusercontent.com/talkops-ai/k8s-autopilot/main/scripts/install.sh | bash
One-Line Shell Installer
Automated Kubeconfig Discovery
20+ Model Providers Supported
Interactive A2UI Web Dashboard