ICP Research Pipeline
A 9-agent orchestration that turns tribal knowledge and CRM data into a validated ICP scoring model with weighted attributes.
The ICP research pipeline chains 9 agents in sequence: context gatherer, tribal knowledge extractor, CRM pattern analyzer, hypothesis generator, evidence researcher, attribute scorer, weight calibrator, model validator, and output formatter. The pipeline starts with what your team already knows (closed-won patterns, gut feel on good accounts) and ends with a quantified model that scores prospects on fit and intent. Each agent writes structured output that the next agent consumes. The full pipeline runs in roughly 25 minutes and produces a scoring model that plugs directly into the icp-builder skill for ongoing use.
Where it shows up:
Related Terms
Methodology
ICP Model
An Ideal Customer Profile definition with firmographic criteria, behavioral signals, and pain hypoth...
Skills Layer
ICP Builder
A multi-session ICP development skill with a 9-agent pipeline that builds, discovers, scores, and de...
Architecture
Upstream/Downstream
The separation of research agents (describe reality) from recommendation agents (interpret and activ...
Skills Layer
Hypothesis Builder
Generates testable pain hypotheses from ICP data, producing structured assumptions that sales conver...