Skip to main content
Glama
dfridkin

clawops

by dfridkin
OverviewSchemaSourceRelated ServersDiscussionsAdmin
TypeScript
Hybrid

Related Servers

Alternatives to clawops

Related Servers

  • Havlio

    kanduit-lab
    F
    license
    Not graded
    quality
    Not graded
    maintenance
    Self-hosted PaaS MCP server providing a write-capable control plane with over 100 tools to deploy, roll back, manage domains, and administer infrastructure, while ensuring agent permissions never exceed human RBAC.
    -

  • infra-mcp

    Phloraxx
    F
    license
    Not graded
    quality
    B
    maintenance
    Securely inspect and operate self-hosted infrastructure including Linux host health, Docker containers and logs, Dokploy deployments, and OCI resources through a narrowly scoped, auditable MCP gateway.
    -

  • InfraClaude

    batu-sonmez
    A
    license
    B
    quality
    D
    maintenance
    An MCP server that enables Claude to manage infrastructure across Kubernetes, Docker, Prometheus, and Terraform through natural language. It provides over 42 specialized tools with a safety-first design, including risk-based command classification and audit logging.
    43
    MIT

  • KubeWhisper

    abdulhusainahk
    A
    license
    Not graded
    quality
    D
    maintenance
    An AI-DevOps MCP server that gives LLMs read-only-by-default access to Kubernetes clusters, Prometheus metrics, and GitHub Actions, enabling natural language queries about infrastructure status and safe write operations with previews.
    MIT

  • InfraPilot

    Gsfrota
    A
    license
    Not graded
    quality
    D
    maintenance
    Enables infrastructure operations including provisioning, configuration, monitoring, compliance auditing, and auto-remediation through natural language, using Terraform and Ansible tools exposed over MCP.
    MIT
Deploy Server
One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

TDQS

A4.1/5.0

Scored across 20 tools

Disambiguation4/5

Most tools target distinct actions on stacks, config, the gateway, or workflows, and the use/don't-use notes are unusually explicit. The main risk is the observability cluster (status, doctor, monitor, logs_tail, workflow_recover), where an agent may need to think before choosing, but the descriptions mostly separate them.

Naming Consistency4/5

All tool names share the clawops_ prefix, use snake_case, and many follow a resource_action shape such as stacks_list, config_get, or gateway_restart. The pattern is weakened by action-only names like up, plan, apply, doctor, and status, but the overall style is predictable and readable.

Tool Count3/5

20 tools is on the heavy side for a single server, especially with two workflow-wrapping tools and multiple observability tools that partly overlap. That said, the lifecycle, config, and diagnostic groups are each justified, so it is not unmanageably bloated.

Completeness4/5

The surface covers stack lifecycle (init/plan/apply/up/destroy), config CRUD and validation, gateway restart, hardening, health checks, logs, and status. Obvious gaps like stack refresh, backup, streaming logs, and gateway update are explicitly deferred to the CLI, which agents can work around by telling the user to run those commands.

Maintenance

ActivityActive
ResponsivenessUnresponsive

Resources