[{"data":1,"prerenderedAt":73},["ShallowReactive",2],{"term-l\u002Flong-horizon-task":3,"related-l\u002Flong-horizon-task":59},{"id":4,"title":5,"acronym":6,"body":7,"category":40,"description":41,"difficulty":42,"extension":43,"letter":44,"meta":45,"navigation":46,"path":47,"related":48,"seo":53,"sitemap":54,"stem":57,"subcategory":6,"__hash__":58},"terms\u002Fterms\u002Fl\u002Flong-horizon-task.md","Long-Horizon Task",null,{"type":8,"value":9,"toc":33},"minimark",[10,15,19,23,26,30],[11,12,14],"h2",{"id":13},"eli5-the-vibe-check","ELI5 - The Vibe Check",[16,17,18],"p",{},"There's a big difference between 'make this button blue' and 'migrate our entire billing system to a new provider.' One is a job an agent nails in a single pass. The other takes hundreds of steps across dozens of files, and the hard part stops being any individual step and becomes simply not losing the plot over that many steps.",[11,20,22],{"id":21},"real-talk","Real Talk",[16,24,25],{},"A long-horizon task is an agentic task requiring many sequential steps, tool calls, or subtask decompositions to complete, where compounding error rates, context management, and plan drift become the dominant failure modes rather than any single step's difficulty. Benchmarks increasingly measure agent capability by the task length at which success rate drops below a threshold like 50 percent, rather than by pass or fail on isolated short tasks.",[11,27,29],{"id":28},"when-youll-hear-this","When You'll Hear This",[16,31,32],{},"\"A button color tweak is a five-second job, a billing system migration is a long-horizon task.\" \u002F \"Most agents fall apart somewhere past step fifty on long-horizon tasks.\"",{"title":34,"searchDepth":35,"depth":35,"links":36},"",2,[37,38,39],{"id":13,"depth":35,"text":14},{"id":21,"depth":35,"text":22},{"id":28,"depth":35,"text":29},"ai","There's a big difference between 'make this button blue' and 'migrate our entire billing system to a new provider.","intermediate","md","l",{},true,"\u002Fterms\u002Fl\u002Flong-horizon-task",[49,50,51,52],"Agent Loop","Context Rot","Plan-and-Execute","Trajectory",{"title":5,"description":41},{"changefreq":55,"priority":56},"weekly",0.7,"terms\u002Fl\u002Flong-horizon-task","x3bnJajjyvUmOkCiQa29qVeCK8nnc4zydKAXCo5F0Gw",[60,63,66,69],{"title":49,"path":61,"acronym":6,"category":40,"difficulty":42,"description":62},"\u002Fterms\u002Fa\u002Fagent-loop","The agent loop is the heartbeat of any AI coding agent: think, act, look, repeat.",{"title":50,"path":64,"acronym":6,"category":40,"difficulty":42,"description":65},"\u002Fterms\u002Fc\u002Fcontext-rot","Context rot is when a long AI conversation has been running so long that the model is confusing old instructions with new ones, forgetting what it already...",{"title":51,"path":67,"acronym":6,"category":40,"difficulty":42,"description":68},"\u002Fterms\u002Fp\u002Fplan-and-execute","Plan-and-execute is writing the full grocery list before you enter the store instead of wandering the aisles improvising item by item.",{"title":52,"path":70,"acronym":6,"category":40,"difficulty":71,"description":72},"\u002Fterms\u002Ft\u002Ftrajectory","advanced","A trajectory is the full replay of everything an agent did to get from your prompt to its final result, not just the destination but the whole road trip: e...",1792528097666]