
incident.io's own plugin: respond to incidents, manage on-call and write the runbooks and skills its investigations draw on.
by incident-io· Cloud & DevOpsUnclaimed · Yours? Claim it
4 images
About
incident.io's own plugin for coding agents. It bundles incident.io's agent skills with its official remote MCP server, so your agent can work with incidents, on-call and postmortems in your incident.io workspace, and help you write the runbooks, architecture docs and skills that incident.io investigations draw on. What its skills cover - Incidents: read and update status, severity, roles, custom fields and updates, and look back at past incidents - On-call: see who is on call, manage overrides and cover requests, and page people and check the page reached them - Postmortems: find, read, edit and review them, checking their claims against what actually happened - Your catalog and architecture: who owns a service, what it depends on, and how your systems are built, deployed and run - Telemetry: query logs, metrics, traces, dashboards and databases for evidence - Extensions: scaffold and register incident.io plugins, write and verify skills, and review the health of your setup How to install it In Claude Code, run /plugin install incident-io@claude-plugins-official, or add incident-io/skills as a marketplace and install incident-io from it. In Claude Desktop it's under Customize, then Plugins. It's also in Cursor's marketplace and the ChatGPT plugin directory for Codex, installs in VS Code with Chat: Install Plugin From Source, and adds to GitHub Copilot CLI from the incident-io/skills marketplace. The MCP server It includes incident.io's remote MCP server at mcp.incident.io/mcp. It asks you to sign in to incident.io the first time a tool is used, and acts as you, seeing what you'd see in the dashboard. The remote MCP server is on incident.io's Team, Pro and Enterprise plans, not the free Basic plan, and is turned on under Settings, then MCP. Good to know - incident.io recommends running these skills on a top model such as Claude Opus; smaller models follow them less reliably. - The doctor skill reports what to fix in your setup and never edits anything. - Status page updates are only drafted; a person reviews and publishes them. - The plugin carries incident.io's guidance, not your content: runbooks and architecture docs stay in your own repositories.
Once it’s added
“Who's on call for the platform team right now?”
“Show me INC-123 with the investigation findings and postmortem.”
“Update INC-123 to resolved.”
“Review the postmortem for INC-456 against what actually happened.”
“Help me get started with incident.io extensions.”
“Create a new plugin, then interview me to write our architecture docs.”
Before you add it
Reads only
Incidents, alerts, schedules and postmortems in your incident.io workspace · Your catalog: who owns a service and what it depends on · Logs, metrics, traces and dashboards from connected observability tools
Makes changes · your agent can ask first
Updates incidents: status, severity, roles and custom fields · Changes on-call: overrides, cover requests and pages · Edits postmortems, and writes architecture docs and skills in your repositories · Scaffolds and registers incident.io plugins
From people who use it
Step by step
Nothing new reaches your agent until a person has looked at it.
No reviews yet. Used incident.io plugin? Say how it went.
Write a reviewRuns on your machine.
Runs on your machine.
/plugin marketplace add anthropics/knowledge-work-plugins /plugin install incident-io@knowledge-work-plugins
Scanned and passed · checked 6 h ago
Nobody has shared how they use incident.io plugin yet.
Plugins people install alongside incident.io plugin.
Scanned on every change, and again when our rules improve.
Tell us and a person on the trust team will look at it. Or email [email protected].
Plugins people install alongside incident.io plugin.
Microsoft's MCP server for Azure: your agent works with Storage, Cosmos DB, AKS, App Service, Key Vault, Monitor and other Azure services.
Cloudflare's skills for Workers, Durable Objects, the Agents SDK and Wrangler, plus its MCP server, in one plugin.
Security scan
Passed
Safe to install in every agent it supports. Claude plugins.