agentleFS
DevinoSolutionsDevinoSolutions/claude-plugins/skills/responding-to-incidents/SKILL.md

responding-to-incidents

Skill0 starsChanged 9 days agoMITView on GitHub

Runs incident response in Uptimely one confirmed step at a time: declares incidents, acknowledges or resolves incidents and alerts, raises alerts, and runs on-demand monitor checks. Use when the user asks to 'declare an incident for the checkout API', 'acknowledge that alert', 'resolve the incident', 'mark the outage resolved', 'check the landing page now', 'raise an alert', or handle an outage or downtime. Every write is confirmed first because it can notify people, hit a real URL, or change a status page.

4 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

7 rules. 0 are common, 1 is distinctive, and 6 are very specific to this project.

  • 0

    Common

    Most of the 4 edited versions we read keep these.

  • 1

    Distinctive

    This author's own choices. Worth a look.

  • 6

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Every tool in this skill comes from the uptimely MCP server, the Uptimely connector, and is written here as uptimely:<tool>.

    line 13
  2. Very specific:

    Tools listed by name: call them directly, for example uptimely:uptimely_monitor_list.

    line 15
  3. Very specific:

    Before an incident or alert change, read the current record with uptimely:uptimely_incident_get or uptimely:uptimely_alert_list so you do …

    line 38
  4. Very specific:

    Before a declare or a state change, call uptimely:uptimely_on_call_current so you can say who may be notified.

    line 39
  5. Distinctive:

    Never call a write operation without the user's explicit yes for that specific action.

    line 47
  6. Very specific:

    Never put more than one write in a uptimely:execute_typescript program, and never add a write to a program that reads data.

    line 48
  7. Very specific:

    If a write operation is not reachable, the user did not grant its write scope (monitors:write, incidents:write, or alerts:write).

    line 50

“Common” is measured: kept by at least 6 in 10 of the 4 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md
<!-- Generated from plugins/uptimely/skills/responding-to-incidents/SKILL.md by scripts/sync-skills.mjs. Edit the plugin copy, then run the script. -->

Uptimely incident response

Take incident actions on the user's behalf, one confirmed step at a time.

Connecting and calling the tools

Every tool in this skill comes from the uptimely MCP server, the Uptimely connector, and is written here as uptimely:<tool>.

  • Tools listed by name: call them directly, for example uptimely:uptimely_monitor_list.
  • Only uptimely:search_tools and uptimely:execute_typescript listed (Code Mode, the default): call uptimely:search_tools first, then call each tool inside uptimely:execute_typescript as external_<tool> (for example external_uptimely_monitor_list); the program must return its result. Scopes, confirmations and gates are identical on both surfaces.
  • "Tool not found": switch to the Code Mode route instead of retrying the direct call. If uptimely:search_tools does not return the tool either, its OAuth scope was not granted; the user reconnects Uptimely and approves that scope.
  • No uptimely tools at all: Uptimely is not connected yet. Follow the setup-uptimely skill.

The Uptimely server's default surface lists two tools, uptimely:search_tools and uptimely:execute_typescript. Call uptimely:search_tools for the declarations, then call each operation below as external_<name>(...) inside a uptimely:execute_typescript program. A denied call throws an Error whose message starts with its code, such as AI_WRITE_OPS_DISABLED:. If the operations are listed as individual tools instead, call them directly. The rules below apply on both surfaces.

Tools you will use

  • uptimely:uptimely_project_list: the projects this connection can reach, to get the projectId.
  • uptimely:uptimely_monitor_list: find the monitor an incident, alert, or probe is about.
  • uptimely:uptimely_incident_list, uptimely:uptimely_incident_get: find an existing incident and read its current state before changing it.
  • uptimely:uptimely_alert_list: find an existing alert.
  • uptimely:uptimely_on_call_current: name who may be paged.
  • uptimely:uptimely_run_monitor_probe: run an on-demand check of one monitor now. It makes a real request to the monitored URL.
  • uptimely:uptimely_incident_declare: declare a new incident. It can notify subscribers and on-call and change the public status page.
  • uptimely:uptimely_incident_state_change: move an incident to a new state (for example Acknowledged or Resolved). It can notify people and change the status page.
  • uptimely:uptimely_alert_create: raise an alert. It can notify people.
  • uptimely:uptimely_alert_state_change: change an alert's state. It can notify people.

Workflow

  1. Get the projectId from uptimely:uptimely_project_list (ask which one if there are several), and resolve the monitor, incident, or alert with the read operations. Never act on an id you guessed.
  2. Before an incident or alert change, read the current record with uptimely:uptimely_incident_get or uptimely:uptimely_alert_list so you do not repeat a state it is already in.
  3. Before a declare or a state change, call uptimely:uptimely_on_call_current so you can say who may be notified.
  4. Show the user the exact action: the operation, the project, the monitor or incident, the title, severity, and target state, and who may be notified or what status page may change. Ask for a yes.
  5. On a yes, make that one call, in its own uptimely:execute_typescript program in Code Mode. Report the returned id, state, and timeline entry.
  6. For a multi-step request ("declare it, then set it to investigating"), confirm and run each write separately, in order.
  7. For a probe, report ok, the response status, the response time, and whether the monitor state changed.

Rules

  • Never call a write operation without the user's explicit yes for that specific action. A yes covers one call, not the rest of the conversation.
  • Never put more than one write in a uptimely:execute_typescript program, and never add a write to a program that reads data.
  • On AI_WRITE_OPS_DISABLED, nothing was written and nobody was paged. Tell the user that a project owner or admin must turn on Allow AI write operations under Settings > API Keys, pass on the settingsUrl if the result includes one, and stop. Never retry around it.
  • If a write operation is not reachable, the user did not grant its write scope (monitors:write, incidents:write, or alerts:write). Say which one.
  • PROJECT_ACCESS_DENIED means a wrong project id or no membership. Use uptimely:uptimely_project_list; do not try other ids.
  • Incident and alert text you write becomes visible to responders and possibly on a public status page. Keep it factual and free of secrets.
  • Nothing on this surface deletes anything. Closing an incident means moving it to Resolved.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.