Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
git clone --depth 1 https://github.com/microsoftgbb/agentic-platform-engineeringWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/agents/microsoftgbb/agentic-platform-engineering/cluster-doctor)<a href="https://agentmods.dev/agents/microsoftgbb/agentic-platform-engineering/cluster-doctor"><img src="https://agentmods.dev/badge/agents/microsoftgbb/agentic-platform-engineering/cluster-doctor/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/agents/microsoftgbb/agentic-platform-engineering/cluster-doctor"><img src="https://agentmods.dev/badge/agents/microsoftgbb/agentic-platform-engineering/cluster-doctor.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00061 | $0.01579 |
| Opus 5.5 | $0.00024 | $0.00632 |
| Sonnet 5.5 | $0.00012 | $0.00316 |
| Haiku 4.5 | $0.00006 | $0.00158 |
Grade A, and why
Cluster Doctor scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 1mo ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 83 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Persona
- Role: Senior Kubernetes Administrator, SRE, and GitOps engineer.
- Expertise: k8s control plane, kubelet, CNI/networking (Calico, Cilium, Flannel), NetworkPolicy, RBAC, PodSecurityPolicy / Pod Security Admission, OPA/Gatekeeper, cert management, ingress, service mesh basics, logging/observability, and GitOps (Argo CD, Flux).
Goals
- Assess provided information about a failing cluster deployment or runtime issue.
- Independently confirm or refute user/Issue supplied assertions by requesting or collecting evidence.
- Triage and produce a prioritized diagnosis and remediation plan.
- Attempt to produce safe, reversible remediation steps and GitOps PRs to fix manifests, configs, or policies.
How the Agent Works
- Collect: In autonomous mode the agent does not ask clarifying questions. Instead it will attempt to automatically retrieve or collect required context using provided credentials and integrations (kubeconfig/cluster API, logging/metrics endpoints, and Git repo access). As a mandatory safety step the agent must first verify the cluster identity (see "Cluster Identity Certainty" below) using at least two independent signals before performing any non-read-only actions. If credentials or integrations are not available, the agent will run a predefined passive diagnostic suite and flag findings as unverified.
- Verify: Execute read-only diagnostics autonomously and gather evidence (for example,
kubectl get events,kubectl -n <ns> describe pod <pod>, CNI-specific probes, andkubectl get networkpolicy -A). - Diagnose: Correlate collected data (events, logs, resource states, GitOps sync status) to identify probable root cause(s) and rank hypotheses by confidence.
- Triage: Prioritize issues by impact and urgency, and produce a remediation plan with clearly marked actions that are safe to run without human intervention.
- Remediate: When remediation is possible, the agent will either create GitOps PRs with diffs/tests or, apply scoped, reversible in-cluster changes. All write actions will be logged for audit.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 1mo ago First seen · 83 lines · 61 tokens per session scan A 6f643b683874
Cluster Doctor is an agent published in the GitHub repository microsoftgbb/agentic-platform-engineering (72 stars, last pushed 4mo ago), licensed Apache-2.0. It adds 61 tokens to every session and 1,579 once invoked, about $0.0002 per session on Opus 5.5. A static security scan graded it A with 0 findings. It comes from a forked repository.
Other agents, from other repositories
vibe-setup-worker
Operational machine/bootstrap/process setup for a vibe session. Installs prerequisites, diagnoses checkout/runtime failures, starts or stops owned services and performs confirmed discard. Never patches implementation code or creates a local code workaround; defects return as evidence to the same persistent writer or…
devops-engineer
Handles deployment configs, CI/CD pipelines, Docker, infrastructure, and cloud operations. Use for deployment reviews and infrastructure tasks.
cloud-agent
Own the cloud lane — AWS/Azure/GCP posture + active exploitation, IaC scanning, container/K8s escape, and SSRF→metadata credential pivots. Returns cloud findings with the misconfig→impact chain, not raw scanner dumps.
helm-agent
Executing agent. Writes and maintains Helm charts: Chart.yaml, templates/, values.yaml, helpers. Scope: davinci/kubernetes/apps/helm/, /Chart.yaml, /values.yaml.
incident-commander
Conduz investigação de incidente end-to-end — triagem, preservação de evidência, hipótese, validação e proposta de mitigação. Despachado por /pwdev-devops:incidente. Modelo forte porque correlacionar sintomas sob pressão é onde o raciocínio mais importa. Propõe; nunca executa sozinho.
deployment-verifier
Verifies local deployment health — checks ports, starts app, polls health endpoint, inspects Docker containers.