[{"data":1,"prerenderedAt":306},["ShallowReactive",2],{"blog-post-blog_en-cursor-automations-agenten-mit-triggers-und-mcp":3},{"id":4,"title":5,"body":6,"cover":290,"date":291,"description":292,"draft":293,"extension":294,"meta":295,"navigation":296,"path":297,"seo":298,"stem":299,"tags":300,"__hash__":305},"blog_en\u002Fen\u002Fblog\u002Fcursor-automations-agenten-mit-triggers-und-mcp.md","Cursor Automations: Always-on Agents with Triggers and Webhooks",{"type":7,"value":8,"toc":285},"minimark",[9,36,41,44,85,92,96,99,137,140,257,260,264,281],[10,11,12,13,17,18,21,22,21,25,21,28,31,32,35],"p",{},"Agentic developer tools are increasingly operated as event-driven systems. With ",[14,15,16],"strong",{},"Automations"," (announced in early March 2026), Cursor extends the agent model with runs triggered by schedules or events from tools such as ",[14,19,20],{},"GitHub",", ",[14,23,24],{},"Slack",[14,26,27],{},"Linear",[14,29,30],{},"PagerDuty",", and ",[14,33,34],{},"webhooks",".",[37,38,40],"h2",{"id":39},"how-automations-work-technically","How Automations Work Technically",[10,42,43],{},"The core is a standardized flow from event to output:",[45,46,47,55,58,64,71,78],"ul",{},[48,49,50,51,54],"li",{},"A ",[14,52,53],{},"trigger"," starts a run based on a schedule or external event",[48,56,57],{},"Schedules and events can be combined in the same workflow (for example nightly + on-incident)",[48,59,60,61],{},"The agent executes inside an isolated ",[14,62,63],{},"cloud sandbox",[48,65,66,67,70],{},"Tool and system access is mediated through configured ",[14,68,69],{},"MCPs"," (Model Context Protocols)",[48,72,73,74,77],{},"Results are published as a ",[14,75,76],{},"pull request",", comment, ticket update, or report",[48,79,80,81,84],{},"Optional ",[14,82,83],{},"state"," via memory\u002Fcontext so recurring runs remain consistent",[10,86,87],{},[88,89],"img",{"alt":90,"src":91},"Diagram: Trigger → Sandbox Run → Output","\u002Fimg\u002Fblog\u002Fcursor-automations-agenten-mit-triggers-und-mcp-diagram.svg",[37,93,95],{"id":94},"enterprise-control-points","Enterprise Control Points",[10,97,98],{},"For trigger-driven agents, three control areas become central:",[45,100,101,107,113,119,125,131],{},[48,102,103,106],{},[14,104,105],{},"Permissions",": least-privilege access to repos, issues, and deployments",[48,108,109,112],{},[14,110,111],{},"Secrets",": separation of runtime secrets from LLM inputs, rotation and scoping",[48,114,115,118],{},[14,116,117],{},"Traceability",": audit logs for prompts, tool calls, and diffs",[48,120,121,124],{},[14,122,123],{},"Quality gates",": reviews, tests, and policy checks before merge\u002Frelease",[48,126,127,130],{},[14,128,129],{},"Data handling",": retention, data residency requirements, and access to internal artifacts",[48,132,133,136],{},[14,134,135],{},"Model routing",": selecting models per automation and central policy for tool calls",[10,138,139],{},"A neutral run contract can capture technical boundaries as a data model:",[141,142,147],"pre",{"className":143,"code":144,"language":145,"meta":146,"style":146},"language-json shiki shiki-themes github-light github-dark","{\n  \"automation\": \"nightly-security-scan\",\n  \"allowedPaths\": [\"src\u002F\", \"infra\u002F\"],\n  \"maxDiffLines\": 600,\n  \"requiresReview\": true,\n  \"outputs\": [\"pull_request\", \"report\"],\n  \"secretsPolicy\": \"no-secrets-in-prompts\"\n}\n","json","",[148,149,150,159,176,196,209,222,240,251],"code",{"__ignoreMap":146},[151,152,155],"span",{"class":153,"line":154},"line",1,[151,156,158],{"class":157},"sVt8B","{\n",[151,160,162,166,169,173],{"class":153,"line":161},2,[151,163,165],{"class":164},"sj4cs","  \"automation\"",[151,167,168],{"class":157},": ",[151,170,172],{"class":171},"sZZnC","\"nightly-security-scan\"",[151,174,175],{"class":157},",\n",[151,177,179,182,185,188,190,193],{"class":153,"line":178},3,[151,180,181],{"class":164},"  \"allowedPaths\"",[151,183,184],{"class":157},": [",[151,186,187],{"class":171},"\"src\u002F\"",[151,189,21],{"class":157},[151,191,192],{"class":171},"\"infra\u002F\"",[151,194,195],{"class":157},"],\n",[151,197,199,202,204,207],{"class":153,"line":198},4,[151,200,201],{"class":164},"  \"maxDiffLines\"",[151,203,168],{"class":157},[151,205,206],{"class":164},"600",[151,208,175],{"class":157},[151,210,212,215,217,220],{"class":153,"line":211},5,[151,213,214],{"class":164},"  \"requiresReview\"",[151,216,168],{"class":157},[151,218,219],{"class":164},"true",[151,221,175],{"class":157},[151,223,225,228,230,233,235,238],{"class":153,"line":224},6,[151,226,227],{"class":164},"  \"outputs\"",[151,229,184],{"class":157},[151,231,232],{"class":171},"\"pull_request\"",[151,234,21],{"class":157},[151,236,237],{"class":171},"\"report\"",[151,239,195],{"class":157},[151,241,243,246,248],{"class":153,"line":242},7,[151,244,245],{"class":164},"  \"secretsPolicy\"",[151,247,168],{"class":157},[151,249,250],{"class":171},"\"no-secrets-in-prompts\"\n",[151,252,254],{"class":153,"line":253},8,[151,255,256],{"class":157},"}\n",[10,258,259],{},"Such contracts matter especially when automations are integrated into change-management processes and outputs must remain reproducible and auditable.",[37,261,263],{"id":262},"why-this-matters","Why This Matters",[10,265,266,267,270,271,21,274,31,277,280],{},"Always-on agents move AI assistance from interactive IDE sessions into ",[14,268,269],{},"operational workflows",". That shift makes ",[14,272,273],{},"access control",[14,275,276],{},"auditability",[14,278,279],{},"reproducible quality gates"," core requirements for production use.",[282,283,284],"style",{},"html pre.shiki code .sVt8B, html code.shiki .sVt8B{--shiki-default:#24292E;--shiki-dark:#E1E4E8}html pre.shiki code .sj4cs, html code.shiki .sj4cs{--shiki-default:#005CC5;--shiki-dark:#79B8FF}html pre.shiki code .sZZnC, html code.shiki .sZZnC{--shiki-default:#032F62;--shiki-dark:#9ECBFF}html .default .shiki span {color: var(--shiki-default);background: var(--shiki-default-bg);font-style: var(--shiki-default-font-style);font-weight: var(--shiki-default-font-weight);text-decoration: var(--shiki-default-text-decoration);}html .shiki span {color: var(--shiki-default);background: var(--shiki-default-bg);font-style: var(--shiki-default-font-style);font-weight: var(--shiki-default-font-weight);text-decoration: var(--shiki-default-text-decoration);}html .dark .shiki span {color: var(--shiki-dark);background: var(--shiki-dark-bg);font-style: var(--shiki-dark-font-style);font-weight: var(--shiki-dark-font-weight);text-decoration: var(--shiki-dark-text-decoration);}html.dark .shiki span {color: var(--shiki-dark);background: var(--shiki-dark-bg);font-style: var(--shiki-dark-font-style);font-weight: var(--shiki-dark-font-weight);text-decoration: var(--shiki-dark-text-decoration);}",{"title":146,"searchDepth":161,"depth":161,"links":286},[287,288,289],{"id":39,"depth":161,"text":40},{"id":94,"depth":161,"text":95},{"id":262,"depth":161,"text":263},"\u002Fimg\u002Fblog\u002Fcursor-automations-agenten-mit-triggers-und-mcp-cover.jpg","2026-04-02","Cursor introduces trigger-driven agent runs that operate in cloud sandboxes and return results back to external systems.",false,"md",{},true,"\u002Fen\u002Fblog\u002Fcursor-automations-agenten-mit-triggers-und-mcp",{"title":5,"description":292},"en\u002Fblog\u002Fcursor-automations-agenten-mit-triggers-und-mcp",[301,302,303,304],"AI","Developer Tools","Automation","Governance","YU-wtrOvCLXyGQfASJz6qFUfH5SR9H3wxvxBclDk_8s",1775680578325]