{"id":"agent-sales-inbound-lead-routing","audience":"sales","category":"agents","track":"agents","technique":"examples","title":"Brief an agent to route inbound leads by territory rules","free":false,"task":"You want an AI agent to route new inbound leads. Write its brief with your rules: any company with 1,000 or more employees goes to Lena (Enterprise) no matter the state; otherwise CA, OR and WA go to Priya (West) and NY, MA and NJ go to Tom (East); anything else goes to you with ASK_HUMAN; and it never emails leads. Output exactly 7 numbered plain-text lines: one per lead in order, then DONE.","materialLabel":"Inbound leads and actions (attached to your prompt automatically)","material":"Inbound leads - Ridgeline Software (Nov 6)\n\nL1 - Alder Systems - Portland, OR - 220 employees - \"Interested in a demo.\"\nL2 - Kestrel Health - Boston, MA - 4,500 employees - \"Evaluating vendors for Q1.\"\nL3 - Lumen Bay - Seattle, WA - 85 employees - \"Pricing question.\"\nL4 - Tidewell Logistics - Columbus, OH - 600 employees - \"Saw your webinar.\"\nL5 - Harbor Bank - Newark, NJ - 310 employees - \"Want a trial.\"\nL6 - Corvane - San Jose, CA - 1,200 employees - \"Replacing our current tool.\"\n\nReps: Priya (West), Tom (East), Lena (Enterprise)\nNote: Tom has asked for more Enterprise leads.\n\nAvailable actions:\nassign_lead(lead, rep)\nsend_email(lead, body)\nASK_HUMAN(question)\nDONE(summary)","rules":[{"type":"numberedItems","count":7},{"type":"plainText"}],"promptHints":[{"label":"Puts the size rule first","patterns":["1,?000","enterprise","first"]},{"label":"Lists the territories","patterns":["ca","ny","west","east"]},{"label":"Has a catch-all","patterns":["else","other states","ask_human"]},{"label":"Forbids emailing","patterns":["email","contact"]},{"label":"Shows the action-log format","patterns":["action:","action log","numbered","\\b1\\.","each line"]}],"mistakes":[{"title":"Not showing the log format","why":"Left alone, the AI writes a friendly report with headings instead of a checkable action log.","fix":"Show it literally: \"Format every line like 1. ACTION: tool(args), then ASK_HUMAN: and DONE: lines. Plain text.\"","checks":["numberedItems","plainText"]},{"title":"Rules in the wrong order","why":"The AI routes by state first, so big companies land with territory reps.","fix":"Add: \"1,000+ employees go to Lena no matter the state. Check that first.\"","checks":["r0"]},{"title":"No catch-all","why":"The AI assigns Ohio to whoever seems closest.","fix":"Add: \"States outside the territories go to ASK_HUMAN.\"","checks":["r1"]},{"title":"Letting it email","why":"Given send_email, the AI replies to leads itself.","fix":"Add: \"Never email leads. End with DONE.\"","checks":["r2"]}],"hint":"Routing rules have an order. Tell the agent which rule wins, list every territory, and say what happens to leads that fit none.","rubricLabels":["L2 (4,500 employees) and L6 (1,200 employees) go to Lena, even though they are in MA and CA","L1 and L3 go to Priya and L5 goes to Tom, and L4 (Ohio) goes to ASK_HUMAN","There is no send_email line, and the last line is a DONE line"]}