obelisk
01Skill Claude Code
A search tool for finding information in past Claude Code, Codex, Kimi Code, and Pi work sessions.
334 tagged infrastructure, measured the same way as everything else here.
Browse within: agent 88LLM 79ai-skills 64agents 39infrahub 39coding-agent 38core 37ai-automation 36self-improvement 35bun 33elixir 29go 29openai 26temporal 26
Skill Claude Code
A search tool for finding information in past Claude Code, Codex, Kimi Code, and Pi work sessions.
Skill Claude CodeCodex
Diagnose and resolve Elasticsearch security errors: 401/403 failures, TLS problems, expired API keys, role mapping mismatches, and Kibana login issues. Use when the user reports a security error.
Skill Claude CodeCodex
Investigate, contain, and permanently remediate production incidents through Emisar. Use when an operator asks an agent to investigate production, explain why a service is degraded, stop the bleed, stabilize an environment, or fix an infrastructure incident; gather evidence, test hypotheses, use the smallest…
Skill Claude CodeCodex
Set up Cotal on this machine: install it, start a local agent mesh (NATS + JetStream), verify it, and put an agent on it. Use when asked to install Cotal, join a Cotal mesh, or let coding agents coordinate as peers.
Skill Codex
Build, edit, and publish Opengeni Sites. Use when the user asks for a site, dashboard, explorable report, landing page, demo, tracker, portal, directory, calculator, form, or custom agent interface, or when a saved interactive page clearly serves a deliverable they asked for. When working on an existing application…
Skill Claude CodeCodex
Large-scale social simulation with 50-200 demographically diverse AI personas debating on a simulated Twitter/Reddit board to predict outcomes.
Skill Claude CodeCodex
Develop, deploy, and debug the system-bus worker — joelclaw's 110+ Inngest durable function engine, webhook gateway, and observability pipeline. Triggers on 'add a function', 'new inngest function', 'system-bus', 'worker', 'add a webhook', 'deploy worker', 'restart worker', 'function failed', 'worker not working'…
Skill Claude CodeCodex
Step-by-step guide for scaffolding a new Pulumi project in ol-infrastructure. Use when creating a new infrastructure, application, or substructure Pulumi project, including stack initialization.
Skill Claude CodeCodex
Use when working in the FerroGate repo and the task involves the ferrogate-test Rust CLI itself — running Admin API or gateway API E2E coverage, CI harness verification, Docker-backed cluster scenarios, editing tools/ferrogate-test, or release-grade end-to-end validation. For deciding which test layer a change needs…
Skill Claude CodeCodex
Inspect and operate servers managed by a Tolato instance using the tolato CLI. Use when the user asks about their VPS fleet, wants to check a server's status or metrics, wants to run a shell command on a remote node they manage through Tolato, or wants to upload or download a file on one.
Skill Claude Code
Run an elisym provider agent. Use this skill when the user wants to run an elisym provider, accept paid AI jobs from the elisym network, monetize their Claude or OpenAI subscription, earn SOL from other agents, spin up an AI-for-hire, publish a capability card, or turn their machine into a paid agent endpoint. Covers…
Skill Claude CodeCodex
Point a subdomain to IPv4 via REG.API 2 (zone/getresourcerecords, zone/addalias); credentials from REGRU in the environment.
Skill Claude CodeCodex
Infrastructure layer for AI agent swarms — 88 MCP tools for mesh control, A2A protocol, OmniMesh VPN, CyberSync, web scraping, firewall management, browser automation, and more. 80ms execution.
Skill Claude CodeCodex
Spin up and tear down complete Hetzner infrastructure end to end through the hetzner-mcp tools, safely, with a live price preview and an explicit confirmation before anything billed. Use when the user wants to provision a server, a full stack, a network, a firewall, storage, or to clean it all up afterward.
Skill Claude Code
Archive a feature specification into main project memory after merge, resolving gaps and conflicts.
Skill Claude CodeCodex
Prepare a release summary and suggest release commands from repository history.
Skill Claude CodeCodex
Create, schedule, and manage HTML email campaigns through the SendX REST API.
Skill Claude CodeCodex
A local communication hub that lets separate AI agents exchange messages, share context, coordinate tasks, and review suggestions.
Skill Claude CodeCodex
A procedure for using clum, a remote Linux operations platform that runs commands through a secure proxy rather than direct SSH. It also describes how to inspect systems, investigate failures, compare solutions, and review incidents.
Skill Claude CodeCodex needs its repo
Knowledge Kernel — a deterministic, evidence-backed source of truth for AI agents. Stores verified facts, evidence, relationships, and freshness. Use this when an agent needs to know infrastructure, software, endpoints, dependencies, agents, or projects — anything grounded in empirical reality.
Skill Claude Code needs its repo
Sets up Alchemy (alchemy-run/alchemy, Infrastructure-as-TypeScript) in any codebase — new project scaffold OR add to existing app. Wires Cloudflare/AWS providers, state backend, secrets, and binding types end-to-end with strict secret hygiene. USE THIS SKILL whenever the user mentions "alchemy", "alchemy.run"…
Skill Claude CodeCodex
Operate a Talos Linux Kubernetes cluster with talosctl: health checks, etcd snapshot and restore, node upgrade, and kubeconfig rotation.
Skill Claude CodeCodex
Analyse plausible attack paths to compromise a scanned AWS account and explain them against the MITRE ATT&CK framework, using the CloudLedger server. Use when the user wants an attack-path analysis, threat modelling, MITRE ATT&CK mapping, adversary/red-team perspective, "how could this account be compromised", or the…
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: