AI Pipeline Management Agents

Use when a user asks “Why is this deal stuck?”, “This deal has been sitting in proposal forever,” or “Rescue my stalled deals.” Diagnoses the blocker behind an open, stale deal and gives the owner a cause-based recovery plan. Not for a single quiet-thread bump, which belongs in No-Reply Rescue, or stage-label validation, which belongs in Deal Stage Reality Check.

Diagnose an outbound campaign from its funnel numbers and produce a weekly status report — selling days remaining, current vs target metrics, the single binding constraint, and the one fix for next week. Use this whenever someone shares outbound activity numbers, asks why a campaign isn't producing meetings, wants a weekly or monthly campaign update or client report, says outbound "isn't working," is deciding whether to kill or extend a campaign, or is reviewing SDR performance. Also use when someone asks for a pipeline or outbound review, or hands over dials, connects, replies, or meeting counts without a specific question — raw activity data almost always means a diagnostic is needed. Reach for this before offering generic outbound advice, because the fix depends entirely on which funnel stage is actually broken.

Identify the economic buyer for in-scope opportunities using only call and email content, never CRM fields; filter voicemails and no-connects before counting engagement, score evidence on the fixed 0-100 rubric, surface buyer gaps and approval gates, flag unsupported close dates, and keep CRM write-back off by default and gated. Use when asked who controls budget, who can approve spend, whether a deal is single-threaded, or for buyer-gap, multithreading, forecast, QBR, or deal-review analysis. Not for inferring buyers from MEDDPICC, contact roles, buyer checkboxes, attendance, or other CRM fields.

Use this when you say: plan my day, run my territory, am I on pace, what do I need to close, prep me for this call, help me follow up, update CRM, rescue this deal, find whitespace, build a bigger deal, or close out my day. It shows where you stand against quota, what needs attention, and the next actions that move you toward bigger deals and quota attainment.

Diagnose where pipeline is leaking across the five GTM handoffs: Marketing to Sales, Inbound to Outbound, SDR to AE, AE to Customer Success, and Customer Success to Sales. Use when someone says leads, pipeline, follow-through, or expansion is dropping and needs a systems-level audit of lost context, un-routed intent, repeated discovery, or missed expansion signals. Produces a prioritized handoff map and fix plan. Not for cold-email deliverability, list cleaning, or a generic demand-generation strategy.

Reconstruct the full multi-channel touch history for target accounts and determine which ordered sequence leads to a first appointment and what happens after a deal is lost. Merge read-only CRM activity, LinkedIn outreach, and cold-email activity by email domain, then report first-touch and last-touch attribution, steps-to-appointment, top sequences, and post-loss reactivation. Trigger when the user asks about lead attribution, multi-touch attribution, which channels drive meetings, winning sequences, steps-to-meeting analysis, or post-loss re-engagement. Not for screening a new list against prior replies.

Use when a sales manager asks for a rep scorecard, 1:1 prep, coaching diagnosis, who on the team needs attention, or a weekly team performance review. Pulls live Salesforce or HubSpot data to produce a weighted Performance Health Score for AEs and SDRs, trend, root-cause diagnosis, and one specific manager action. Not for generic sales coaching without live rep data or a named review target.

Use when a complex opportunity needs a MEDDPICC-style evidence audit from calls, emails, and CRM context. It maps what the buyer has actually confirmed, exposes contradictions, and produces the next best evidence-gathering move. Not for scoring a deal from rep opinion alone.

Use when the user wants what is known about one account across multiple meetings: "what do we know about account X across our calls", "what patterns keep showing up with this prospect", "how has this deal evolved over our meetings". Weighs repeated signal over one-off comments and reads momentum honestly. For one meeting's note or follow-up, use the post-meeting skills; for judging the deal, Deal Risk Pressure-Test.

Use when the user asks what to work on this morning or wants their daily brief: "what should I do today", "give me my brief", one prioritized view of must-reply threads, stale-risk deals, CRM gaps, and today's top actions. Not for prepping one call (Pre-Call Prep), one deal's health (Deal Risk Pressure-Test), or drafting messages.

Use when the user wants one deal honestly judged: "pressure-test this deal", "am I being too optimistic about Acme", "what's actually weak in this opportunity". Grades urgency, pain, champion, multi-threading, decision process, proof, and next step against evidence. For a neutral summary before a reply use Deal Context Builder; pipeline-wide auditing lives in Pipeline Truth.