Create Automation (Interactive)

Use this skill when the user explicitly wants to make, build, set up, or create a new Cursor Automation — for example "create a Cursor automation", "open the Automations editor with this draft", or "set up a scheduled Cursor agent".

Disambiguation. "Automation" in a user workspace can mean many things (.github/workflows, CI pipelines, scheduled jobs, scripts, dbt, browser automation, shell scripts, workflow engines). Do not assume generic phrases like "automate this", "help me automate my deploys", or "make an automation" mean Cursor Automation. Route to the named surface when the user mentions one, use normal repo/product exploration when the context points elsewhere, or ask a short clarifying question when the target surface is ambiguous.

Execution spine (every run)

1. Finish-path check. Verify the in-app Automations editor handoff is available. If not, immediately say "Please use this skill in the Agents Window." and stop.
2. Capture intent + proactive integration discovery. If the prompt is missing 2+ of {what kicks it off, what should happen, what outcome}, send one short message asking for a 1-2 sentence description. Proactively run discovery for any integration the user named or implied.
3. Completeness gates. Work through trigger, tool, prompt, naming, and PCD checks. Do not jump to a summary while required fields are still unknown.
4. One consolidated question. Ask inline for whatever integration discovery couldn't resolve.
5. Draft table → approval → finish handoff. Show a compact Markdown table recapping name / description, trigger, tools, instructions, resolved settings, and "to finish in editor". User approves, then open the Automations editor.

House rules

• Plain language only. Never show MCP / tool / proto names, enum values, or raw CLI output in user-visible chat.
• No YAML in finalization. The draft table = compact Markdown only.
• No automatic fallbacks. Never submit, open a URL, paste a browser prefill link, or switch buckets.
• Creation-only. This skill prepares new automations only. Do not list, get, inspect, update, or search existing Cursor Automations from chat.

Trigger categories

| Trigger | Description | |---------|-------------| | On a schedule | Cron-based recurring runs | | On a GitHub / GitLab event | PR opened, pushed, merged, commented, label change, branch push, CI completed | | On a Slack event | New message, reaction added, channel created | | On a Linear event | Issue created, status changed, end of cycle | | On a PagerDuty incident | Incident triggered, acknowledged, resolved, any | | On a Sentry issue | Issue created, resolved, assigned, archived, unresolved, any | | On an incoming HTTP webhook | Webhook trigger |

Available tools

| Label | Purpose | |-------|---------| | Comment on PRs | Post comments on pull requests | | Post to Slack | Send messages to Slack channels | | Read Slack | Read messages from Slack channels | | Request reviewers | Assign PR reviewers | | Manage check runs | Create/update CI check runs | | Use MCP server | Call tools from connected MCP servers |

Authoring funnel

Work in Automations UI order: trigger → tools → prompt → name/description → draft table.

1. Trigger: Pick from the trigger categories above, resolve picker-backed values via integration discovery.
2. Tools: Select which tools the automation needs.
3. Prompt + name: Ask "What should the agent do when trigger]?" Suggest a name + 1-2 sentence description.
4. Draft table: Recap as a compact Markdown table. Get approval.
5. Finish: Open the Automations editor with the reviewed draft.

Draft table format

| Draft field | What will open in the editor | |-------------|------------------------------| | Name / description | Short plain-language value | | Trigger | What starts the automation | | Tools | Enabled capabilities | | Instructions | The prompt behavior, summarized | | Resolved settings | Repo / branch, Slack channel, service / project, schedule | | To finish in editor | Settings deferred to the Automations UI; write "None" if nothing is deferred |