Client agent
It turns meetings and delivery notes into decisions, actions, updates, health checks, and proof requests inside one client workspace. Client-facing work still waits for review.
Summary
Runs
Client onboarding, meeting capture, delivery drafts, project updates, reports, health checks, proof requests, and case-study drafts
Reads
One client workspace, meetings, goals, reports, decisions, offers, and approved templates
Writes
Briefs, meeting logs, decisions, actions, checklists, reports, risk lists, testimonial requests, and draft case studies
Guardrail
One client at a time. No sibling client reads. No client contact without a person
Where it helps
The first agent should have a clear job, known inputs, and a review path the team can keep using.
What it can run
These capabilities come from the public Conversion Skills patterns. The names under each card are the documented skills the build can draw from.
The client skill can onboard a client, create the workspace, pull a strategist brief, log a meeting, draft a deliverable, build a report, run a health check, or archive a client.
It can turn transcripts or notes into a filed meeting record, route decisions, extract actions, append metric rows when numbers are mentioned, and log the work to the daily note.
It can build a dated onboarding checklist with milestones, owners, due dates, and a first-value success measure.
It can refresh a project brief, log progress, append ledger rows for moved metrics, surface blockers, and hand off published work as a status update.
It can scan accounts or client goals for stale metrics, silence, missed milestones, open escalations, renewal risk, and recommended save plays.
It can draft testimonial requests, store received testimonials with consent status, and draft a case study only when a ledger row supports the result.
How it works
Step 01
The agent works only inside the active client workspace. If the client is ambiguous, it asks or files the note for a person instead of guessing.
Step 02
Meetings, decisions, actions, metrics, deliverables, reports, and proof each go to their correct home so the next run can find them.
Step 03
The result says what changed, what is waiting, who owns the next move, and what a person should review before anything client-facing leaves the business.
Human gates
The agent prepares work and keeps the trail clean. People still approve the moves that can affect buyers, customers, money, permissions, or public claims.
Outputs
The useful output is not a magic answer. It is a reviewable artifact, a clearer next move, and a source trail for anything factual.
Firewalled client workspaces and one-screen briefs
Meeting records with routed decisions, metrics, and actions
Onboarding plans with owners, dates, and success measures
Project updates with blockers and next steps
Weekly or monthly client reports
Client health checks, proof requests, and draft case studies
Related agents
Most useful AI systems start with one agent, then connect the next repeated job once the first one earns trust.
Public standard
Conversion Skills is the public reference for these patterns: plain files, documented skills, proof-aware outputs, and review gates. The agent pages show how that public method turns into useful business systems.