Hire an AI pilot customer promise reviewer

Everything it can hand back

Your task sets the actual outputs. The specialist picks the methods that cover them.

• A register with one row per promise, listing what was promised, who accepted it, the owner, the due date, the source, the status and the delivery proof
• A list of promises that conflict, or go beyond your product limits, with both sources shown
• A draft update for each customer, covering only what changed
• A handover note when an owner changes
• A short record for the next review

How it does the job well

• Check the current policy and the customer's order or account facts before replying.
• Answer the actual question in the first two lines, then explain.
• Match the customer's tone. Acknowledge the problem once, without grovelling.
• Hand refunds outside policy, legal threats and safety issues to a person.

Mistakes it avoids

• Counting an enthusiastic reply as a commitment.
• Telling a customer a promise is delivered when only an internal ticket says so.
• Keeping customer personal details beyond what the record needs.
• Calling installation a running service. Nothing runs unless the user starts it.
• Promising refunds, dates or fixes the policy does not allow.
• Copy-paste replies that ignore what the customer said.

What a good result looks like

An example, not your task: A customer asks "did we ship what you promised in March?" and the answer is spread across calls and emails.

• Promise register. One row per promise with its ID, what was promised, who accepted it, owner, due date, source, status and delivery proof.
• What changed. Each promise whose status, owner or due date moved since the last review, and why. A first run says first baseline.
• Conflicts and limits. Promises that clash with each other or go beyond your approved product limits, with both sources shown side by side.
• Customer update drafts. A short draft for each customer covering only what changed. It names no feature or date you have not approved. Marked not sent.
• Handover note. Only when an owner changes. What is promised, what is due, where proof lives and who is proposed. A proposed owner has not accepted.
• Record for next review. A small block to paste next time with pilot ID, promise IDs, accepted owner, due date, delivery source and update version.

How the work is checked

• No promise is accepted on a request or a sales line alone, and each has a source.
• No promise is delivered without a dated delivery record.
• No draft offers a feature or date that is not approved, or says unproved work is done.
• Every conflict shows both sources.
• States which checks passed, failed or could not run, and names missing inputs.

These checks have not been run against a model yet. Your AI uses them to check its own work.

Capabilities

• Reads the last record. Uses the saved record or sets a first baseline. The record holds the pilot ID, promise IDs, accepted owner, due date, delivery source and customer update version.
• Checks the trigger. An accepted promise changed, or proof of delivery arrived. If neither, it says so and stops.
• Lists the promises. The Executive action register keeper pulls each explicit commitment with its source, owner and due date. A request or a sales line is listed as asked, not accepted.
• Checks limits and conflicts. Compares each promise with your approved product limits and with the others. A promise beyond your limits, or two that clash, is flagged with both sources.
• Sets each status. Marks each promise asked, accepted, delivered, claimed not proved or blocked. Delivered needs a dated record.
• Drafts changed-only updates. The Stakeholder update writer drafts a short update for each customer that covers only what changed. It names no feature or date you have not approved.
• Hands over when an owner changes. The Handover note writer lists what is promised, what is due and where proof lives. An owner who has not accepted stays marked as proposed.

How the role works

1. Checks your task, context and available tools.
2. Picks the methods needed and prepares a work plan.
3. Carries out supported tasks, with each method's limits and review gates.
4. Checks the real deliverables and corrects supported faults.
5. Hands back files, evidence and unresolved gaps.

Your specialist's working setup

Download the base specialist without a task form or account. README.md and SETUP.md explain how to use the folder. SOUL.md defines its scope. INTAKE.md asks missing questions inside your AI. PLAYBOOK.md, SKILLS.md, BEST_PRACTICES.md, CHECKS.md and HANDOFF.md guide the work and its checked deliverables.

Complete associated skills, supporting files and licences are included. Choose only the methods needed for the actual task. Keep context, working files and accepted feedback in your AI project.

Use this specialist through MCP or the API

Call prepare_specialist with role_id pilot-customer-promise-review-assistant only. No task is required to load it. Your AI asks what job you need done, what you want back and which relevant context it can use.

For task-led hiring, call find_specialist, explain the suggested specialists and let the user accept or change the team.

Complete ZIP: Download base specialist. API: GET /api/v1/specialists/pilot-customer-promise-review-assistant/pack. The host AI supplies its own model, tools and execution; a pack does not connect accounts or run a task.

Inputs and helpful context

The specialist starts from the task brief. Only active methods require their inputs.

For the primary method, when used

• The pilot agreement or order form, and any written scope (from your contract folder)
• Emails or messages where promises were made
• Notes or transcripts of customer calls, only where the customer agreed to them being kept
• Proof of delivery, such as release notes, a shipped link or a delivery email, each with its date
• The owner and due date you agreed inside your team for each promise
• Your approved product limits, which means what you can and cannot promise (from your product lead)
• The last review's record (optional, needed to see what changed)

Helpful context (optional)

• The customer's own written success criteria for the pilot.
• A named person who can accept promises for your team.
• Your product roadmap with what is approved.

Tools and access

• A chat AI that reads uploaded files: Claude, ChatGPT and Gemini can read your agreement, messages and the pack. In plain chat you save the record yourself.
• Your contract and release notes folders: The agreement gives the written scope. Release notes give dated proof of delivery.
• Any good chat AI that reads files, such as Claude, ChatGPT or Gemini: It needs agreements, notes and delivery records at once.
• Any good chat AI: A one-message note works in any strong chat AI.

Functions and deliverables

• Pilot customer promise reviewer: Track promises made to pilot customers. Check owners and proof of delivery, then draft honest updates.
  • A register with one row per promise, listing what was promised, who accepted it, the owner, the due date, the source, the status and the delivery proof
  • A list of promises that conflict, or go beyond your product limits, with both sources shown
  • A draft update for each customer, covering only what changed
  • A handover note when an owner changes
  • A short record for the next review
• Pilot customer commitment tracker: Share pilot deals, call notes and delivery records. Get a list of agreed promises, owners and dates, plus customer update drafts.
  • A promise register: each accepted promise with source, owner, due date, dependency and proof needed
  • A list of requests and sales statements not yet accepted
  • Clashes between promises, and promises with no owner or date
  • Update drafts for each customer showing only what changed
  • A summary and one next step
• Handover note writer: Turn a brain dump into a handover note. The person covering your work sees what breaks first and what is due, and can act on day one.
  • Time bombs first, each with a date and an action
  • What is due, live work with next steps, and where things live
  • The people and how to deal with them, what to leave alone, how to reach you
• Delivery update flow builder: Share shipping data and order-status contacts. Get a date rule, four updates, a tracking-page plan and a reply for late parcels.
  • Your WISMO rate and delivery accuracy, each rated against a limit
  • A delivery date rule that pads your real times
  • A four-message update flow with timing and copy
  • A tracking page plan and a script for delayed parcels
• Brand profile builder: Share your website. Get a reusable profile covering your products, customers, voice and approved claims, with sources.
  • A folder of eight plain files, from a start file for your AI to a claims list and a change log
  • A tag on every fact, and every claim marked approved, pending or banned
  • A two-sentence diagnosis of your biggest gap, and a list of what's missing and who can answer it
  • A short slice for your next step, if you name one

Tasks you can use it for

• A customer asks "did we ship what you promised in March?" and the answer is spread across calls and emails.
• A new account manager takes over a pilot and needs to know what was promised.
• A feature just shipped and you want to close the matching promises with proof.
• Two customers were promised the same thing for different dates, or a promise goes beyond what the product can do.

Limits to check before starting

• The promise was never written down or recorded anywhere. Write it down first. This tracks promises that have a source.
• You only need to pass work to a new owner. Use the Handover note writer.
• You want to reply to a customer's support message. Use the Customer inbox reply drafter.
• You want to track actions from your own team meetings. Use the Action item tracker keeper.
• You want customers messaged for you. It drafts and stops.

Best way to use this role

Give it one clear outcome, the actual files and a definition of done. Keep the work in one AI project so it can reuse checked context and earlier files.

• Add the agent with the pieces it uses, the Track promises made to pilot customers workflow and the Handover note writer. If your AI cannot open them, paste both in with the agent.
• Attach or paste your agreement, the messages and call notes, the delivery proof and your product limits. If this is a repeat, paste last time's record. Take out personal details your AI does not need.
• Read the conflicts and the promises beyond your limits first. Settle those with your product lead before anything goes to a customer.
• Check that every promise marked delivered has a dated proof. A promise with none moves to "claimed, not proved".
• Edit the update drafts and send them yourself. In plain chat it cannot save files, so keep the record it hands back.

Read it as data

Read the complete role setup as JSON. Connect your AI to the AI Blocks MCP library and read aiblocks://roles/pilot-customer-promise-review-assistant for its role instructions and pinned sources.

This is a prepared AI role. No task has run. Your AI carries out authorized work using its available tools and reports the actual deliverables and checks.