Claude incident.io connectorWhat Claude can do in your incident.io account.
The Claude incident.io connector exposes 35 tools. 27 read alerts, incidents, on-call and catalog data, 6 create or change something, and 2 have no established read or write verdict. Here is what Claude can declare for you and what depends on your incident.io products.
Verified Trustpilot reviews · AI, automation & growth agency
Why connect incident.io to Claude
incident.io is where your alerts, incidents and on-call rotations live. Connected to Claude, it answers in plain English: who is on call for the platform team, how many high-severity incidents hit this quarter, which alerts turn into real incidents and which are just noise. Claude gathers the numbers, cross-checks them and hands you a summary, with no dashboard to open.
Measure instead of guessing. incident_stats rolls up incident counts and workload across more than ten dimensions, and alert_stats counts alerts alongside the workload of the incidents linked to them.
Run a response without leaving the chat. incident_create declares an incident, incident_update moves its status or severity along, and escalation_respond acknowledges or declines a page.
Prep an ops review that holds up. analysis_start loads guided playbooks, your organization's configuration and a report template, for a structured operational analysis.
What Claude does not do here: it never receives alerts on its own. No tool fires when a page goes out at 3am; you are the one asking. It acts with the permissions of whoever approved the connection, no more. And several tools on the sheet only work if your account runs the matching incident.io product. For an automatic reaction to an event, you need an automation tool, a different trade; the Integrations hub covers that side. For a project management connector, see the Claude asana connector.
Vocabulary in one minute
Five words so the rest reads without friction.
- Connector
- The link you set up once between Claude and your incident.io account, so Claude can work in it while it answers you.
- Tool
- One named action the connector opens to Claude. Claude decides which to call by itself, and the directory sheet lists each one by name.
- Authorization
- incident.io's own sign-in screen, where you give Claude the access it will use afterwards. Granted once per person, withdrawn the same way.
- Approval
- The confirmation Claude waits for before it completes an action that would change your account, shown in the chat when it matters.
- MCP
- The shared standard behind connectors, which is what lets an assistant like Claude talk to an outside service such as incident.io.
Connect incident.io to Claude
- 01
Find incident.io in the list
In Claude's settings, open Customize, then Connectors, and search for incident.io. On a Team or Enterprise workspace, an Owner or Primary Owner enables the connector for the organization first; each member then connects on their own.
- 02
Sign in to incident.io
Click Connect on the incident.io row and sign in through the window incident.io opens itself. Pick the account carefully: Claude will see and act under that name.
- 03
Approve the authorization screen
Read incident.io's authorization screen before accepting: it sets what the access covers, not Claude. If the link breaks, use Disconnect and connect again; you can also revoke access from incident.io.
The 35 tools of the Claude incident.io connector
incident.io gives Claude 35 tools: 27 that read your account, 6 that change something in it, and 2 no official source describes.
Twenty-seven lookup tools, six that act, and two AI agents whose effect is unclear. Names stay exactly as Claude shows them.
- 27 read
- 6 write
- 2 not documented
What Claude reads (27)
27 toolsTwenty-seven tools that look at alerts, incidents, on-call, teams and the catalog without touching them.
alert_list
Browses and filters the alerts incident.io has received. Claude uses it to isolate a burst of alerts or find the one that came right before an incident.
alert_show
Opens one alert in full, along with the incidents linked to it. It is the zoom that follows a list.
alert_source_list
Lists the alert sources configured in your account: the monitoring tools that feed incident.io.
alert_stats
Counts alerts and attaches the workload of the incidents they triggered. Claude can then tell alerts with real incident work behind them from the ones that led nowhere.
analysis_start
Loads a full analysis workspace: playbooks, your organization's configuration and a report template, the base for a structured operational analysis.
ask_telemetry
Hands your question to an AI agent that queries logs, metrics, traces and dashboards, then brings the answer back into the conversation.
catalog_entry_list
Walks through the entries of one catalog type, for example every service or every team described in incident.io.
catalog_entry_show
Shows one catalog entry with all its attributes. Claude finds the owner, dependencies or any other detail your team keeps there.
catalog_type_list
Lists the catalog types defined in your account, such as Service or Team. It is the entry point before browsing the entries themselves.
escalation_list
Searches escalations, meaning the pages sent to on-call responders. Reserved for accounts on the On-call product.
escalation_path_list
Lists the configured escalation paths: the sequences that decide who gets paged, and in what order.
escalation_path_show
Shows, level by level, who would be paged if this escalation path fired right now.
escalation_show
Details one escalation with its transition history: who was paged, when, and what happened next.
escalation_stats
Counts pages by escalation path, priority and time of day. A way to measure how heavy on-call really is.
follow_up_list
Lists the follow-ups opened after incidents, the tasks agreed in a post-mortem to stop a repeat.
follow_up_stats
Totals follow-ups by owner, status and other criteria. Useful to see where things get stuck.
incident_list
Searches and filters incidents, with an option to sort by workload. It sits between the stats and the detail of a single incident.
incident_show
Displays one incident in full and, on request, the investigation findings, the post-mortem and the history of status posts.
incident_stats
Aggregates incident counts and workload across more than ten dimensions. Claude turns those totals into a readable summary instead of handing you a raw table of numbers.
incident_update_list
Retraces the full history of status posts published during an incident, in chronological order.
investigation_sync
Downloads a complete investigation as a single file, which incident.io intends for analysis by a language model. Claude can then work through the whole investigation in one go rather than piece by piece.
resource_list
Shows everything available to read: organization configuration, analysis playbooks and telemetry data sources.
resource_show
Reads one resource: organization configuration, an analysis playbook or a telemetry data source. It is the step after the inventory of resources, once Claude knows what exists.
schedule_list
Lists your account's on-call schedules. Claude uses it to know which rotation to query next.
schedule_show
Details one on-call schedule with its current and upcoming shifts.
team_list
Returns the list of teams incident.io knows about. It is the starting point for scoping an analysis to a clear perimeter.
team_show
Presents one team with the escalation paths, alert sources and schedules it owns.
What Claude changes (6)
6 toolsSix tools that act on your account. No source describes a specific confirmation, so the general rule below applies.
escalation_respond
Approval: see the ruleAcknowledges or declines an on-call page from the conversation, without opening the incident.io app.
feedback
Approval: see the ruleSends incident.io feedback about its tools. According to the publisher, the assistant offers this feedback after a task, and it only goes out once you approve it.
follow_up_create
Approval: see the ruleCreates a follow-up attached to an incident, so a fix agreed in a debrief does not get lost.
follow_up_update
Approval: see the ruleEdits a follow-up that already exists on an incident, so the task list stays in line with what the team actually did.
incident_create
Approval: see the ruleDeclares a new incident in incident.io, straight from the conversation.
incident_update
Approval: see the ruleChanges an incident's fields: status, severity, roles or custom fields.
Effect not established (2)
2 toolsincident.io describes these two AI agents, but no source says whether they only read or also change your account.
ask
incident.io presents this tool as an AI agent for on-call questions, schedule management and general queries. No source says, however, whether it only reads or also acts on your account.
ask_incident
According to incident.io, this is an AI agent for incident investigation and management actions. No official text specifies whether it only reads or takes action in your account.
What Claude checks with you first
By default, Claude stops and asks before any action it takes on an account on your behalf. With incident.io that concerns six tools, and possibly two AI agents.
On a Team or Enterprise workspace, owners decide whether members can let some actions through without being asked again. They can also cap what a connector does for the whole organization, for instance keep reading open and block writing, and nobody overrides that from their own account. On the incident.io side, the connection acts as the person who approved it: it sees and does only what that person can. The publisher even recommends connecting a dedicated service account.
Which Claude plans include incident.io
None of the 819 directory sheets shows plan availability. Connector by connector, that information is not published anywhere.
The general rule is public: remote connectors are open to all users on Claude, Cowork, Claude Desktop and mobile. On Team and Enterprise, an Owner or Primary Owner enables the connector for the organization before members connect. For incident.io's current status on your account, check its sheet in the official directory. incident.io also limits its server to some of its own tiers, which is settled on their side.
Where the incident.io connector stops
A connector is not an automation. Claude calls these tools while it answers you, so nothing fires when an alert comes in.
Only the publisher documents this connector: neither Claude's help center nor Anthropic's developer docs have a page on it, and the limits below come from incident.io's documentation. The tool list is an observed floor, since an admin can open actions no public sheet shows. The partner badge is not a security audit, and Anthropic notes on every sheet that it does not choose a publisher's tools or vouch for how they behave. incident.io in fact documents far more tools than the 35 on the sheet.
Need help connecting incident.io to Claude?
A person reads every message.
