abstract agent · marketing · 36 processes · plugin 1.0.0

Account monitor

account-monitor v1

Answers questions about one named account. It reads that account's product usage, its support history, its contract position and where it stands in the renewal cycle, and it comes back within minutes with whether the account has reached the milestones that count, whether a risk signal describes something real, and whether the account is in good enough standing to be approached at all. Every answer arrives with the evidence under it: which system was read, what that system said, and when the reading was current.

The abstract agent

What it is granted, and what it hands to a person

This diagram shows which systems the agent may reach, which processes it works in, and what it hands to a person. Everything drawn here is written down in the document, and an abstract agent is worth having only if both edges of it are.

Answers the roll-call that opens a run about a customer, saying where the account stands before anyone plans what to do about it. Contributes to collect-and-report when a run asks every source the same question, and keeps each line attributed to the system it came from. Sits in the briefing that starts an onboarding, a churn-risk run or a win-back, so every agent that acts hears the same account picture at the same time. The rest of the time it works alone against a queue of questions, which is why the record it leaves with each answer matters as much as the answer.

The plugin

What is in the package

7 skills and the systems they need, written to version 1.1.0 of the Agent Plugins standard. Nothing in these files names a product. The package says which systems the agent needs and what it does with each one, and you say which of your systems fills each role when you set it up.

Download the package View on GitHub

account-monitor · 12 files · 1.0.0

com.agentcatalog.agent/NOT.mdWhat the agent does not do, and when it asks a person.

# Where this abstract agent stops

Does not contact the customer and does not send anything to them. Does
not decide what to do about what it finds; it answers the question and
the agents that act read the answer. Does not choose which messages a
customer receives or run a standing program for them, which is the
lifecycle-manager agent. Does not ask a customer to do anything in
public, which is the advocacy-manager agent, and that agent asks this
one whether the account is in standing before it makes the ask. Does not
score or route an inbound lead, which is the lead-scorer agent. Does not
measure a campaign or a published piece, which is the analytics agent,
and does not own the department's report for a period, which is the
reporting-manager agent; this agent answers about one named account.
Does not change the CRM, the contract, or anything else it reads. Does
not negotiate a renewal or make a commercial offer, which stays with the
people who own the account.

## What always goes to a person

- a risk signal severe enough that waiting for the next scheduled check
  would waste the warning
- two systems that disagree about the same account beyond the tolerance
  the process sets
- a question it cannot answer because a system it needs is unavailable
  or its data is out of date, raised before the answer is due
- a contractual or legal matter it finds while reading, such as a
  dispute, a breach notice or an account in collections, because whether
  that account may be approached at all is a person's decision

com.agentcatalog.agent/RECORDS.mdThe records the agent must leave behind.

# The records this abstract agent must leave

This is the contract. An agent implementing `account-monitor` leaves the
following behind, on every run.

For each question it answers, it records the account, what was asked,
the answer, the date, and every system it read with the value it found
and when that value was current. A milestone answer names the milestone
list at its version and says which of those milestones the account has
met. A risk signal is recorded whichever way it was judged, with the
reading that settled it. Anything it could not read is recorded as
unread instead of being left out of the answer.

An agent that leaves less than this does not implement
`account-monitor`, whatever else it does well.

com.agentcatalog.agent/SERVERS.mdThe systems the agent needs, by role rather than by product.

# The systems this abstract agent needs

Five systems, named by role rather than by product. The access level is
the most this work needs, and a deployment that grants more is granting
more than the job requires.

| Role | Access | What the access is for |
|---|---|---|
| product usage data | read |  |
| the support system | read |  |
| the CRM | read |  |
| contract and renewal records | read |  |
| the account standing record | write | Each answer lands dated, with the evidence under it. |

Whoever builds an agent from this plugin writes the `mcp.json` that
points these roles at their own systems. Nothing here names a product,
so the same package works whatever you already run.

plugin.jsonThe manifest. It lists the skills and the systems the agent needs.

{
  "$schema": "https://agent-plugins.org/schemas/1.1.0/plugin.schema.json",
  "name": "account-monitor",
  "version": "1.0.0",
  "description": "The account-monitor abstract agent. Answers questions about one named account.",
  "author": {
    "name": "AgentCatalog",
    "url": "https://agentcatalog.com"
  },
  "homepage": "https://agentcatalog.com/abstract-agents/account-monitor",
  "repository": "https://github.com/jeffrschneider/agentcatalog",
  "license": "CC-BY-4.0",
  "keywords": [
    "abstract-agent",
    "reference",
    "marketing",
    "onboarding-email-program",
    "churn-risk-play",
    "account-based-play",
    "abstract"
  ],
  "extensions": {
    "com.agentcatalog.agent": {
      "agent": "account-monitor",
      "agentVersion": 1,
      "department": "marketing",
      "abstract": true,
      "works": "Answers the roll-call that opens a run about a customer, saying where the account stands before anyone plans what to do about it. Contributes to collect-and-report when a run asks every source the same question, and keeps each line attributed to the system it came from. Sits in the briefing that starts an onboarding, a churn-risk run or a win-back, so every agent that acts hears the same account picture at the same time. The rest of the time it works alone against a queue of questions, which is why the record it leaves with each answer matters as much as the answer.",
      "servers": [
        {
          "name": "product usage data",
          "role": "product-usage-data",
          "access": "read"
        },
        {
          "name": "the support system",
          "role": "support-system",
          "access": "read"
        },
        {
          "name": "the CRM",
          "role": "crm",
          "access": "read"
        },
        {
          "name": "contract and renewal records",
          "role": "contract-and-renewal-records",
          "access": "read"
        },
        {
          "name": "the account standing record",
          "role": "account-standing-record",
          "access": "write",
          "needs": "Each answer lands dated, with the evidence under it."
        }
      ],
      "records": [
        "For each question it answers, it records the account, what was asked, the answer, the date, and every system it read with the value it found and when that value was current.",
        "A milestone answer names the milestone list at its version and says which of those milestones the account has met.",
        "A risk signal is recorded whichever way it was judged, with the reading that settled it.",
        "Anything it could not read is recorded as unread instead of being left out of the answer."
      ],
      "escalates": [
        "a risk signal severe enough that waiting for the next scheduled check would waste the warning",
        "two systems that disagree about the same account beyond the tolerance the process sets",
        "a question it cannot answer because a system it needs is unavailable or its data is out of date, raised before the answer is due",
        "a contractual or legal matter it finds while reading, such as a dispute, a breach notice or an account in collections, because whether that account may be approached at all is a person's decision"
      ],
      "not": [
        "Does not contact the customer and does not send anything to them.",
        "Does not decide what to do about what it finds; it answers the question and the agents that act read the answer.",
        "Does not choose which messages a customer receives or run a standing program for them, which is the lifecycle-manager agent.",
        "Does not ask a customer to do anything in public, which is the advocacy-manager agent, and that agent asks this one whether the account is in standing before it makes the ask.",
        "Does not score or route an inbound lead, which is the lead-scorer agent.",
        "Does not measure a campaign or a published piece, which is the analytics agent, and does not own the department's report for a period, which is the reporting-manager agent; this agent answers about one named account.",
        "Does not change the CRM, the contract, or anything else it reads.",
        "Does not negotiate a renewal or make a commercial offer, which stays with the people who own the account."
      ],
      "processes": [
        {
          "process": "ref/mkt/account-based-play",
          "activities": [
            "Read the Accounts",
            "Watch Engagement by Account"
          ]
        },
        {
          "process": "ref/mkt/advisory-board",
          "activities": [
            "Settle the Membership"
          ]
        },
        {
          "process": "ref/mkt/award-entry",
          "activities": [
            "Get the Customer's Agreement"
          ]
        },
        {
          "process": "ref/mkt/churn-risk-play",
          "activities": [
            "Confirm the Signal",
            "Watch What the Account Does",
            "Record the Outcome"
          ]
        },
        {
          "process": "ref/mkt/customer-case-study",
          "activities": [
            "Check Where the Account Stands",
            "Verify the Customer's Results"
          ]
        },
        {
          "process": "ref/mkt/customer-newsletter",
          "activities": [
            "Check Where the Account Stands"
          ]
        },
        {
          "process": "ref/mkt/customer-reference",
          "activities": [
            "Check Each Account's Standing"
          ]
        },
        {
          "process": "ref/mkt/customer-reviews",
          "activities": [
            "Choose the Customers to Ask"
          ]
        },
        {
          "process": "ref/mkt/display-retargeting",
          "activities": [
            "Build the Suppression List"
          ]
        },
        {
          "process": "ref/mkt/enable-channel-partner",
          "activities": [
            "Check What the Partner Holds",
            "Watch What the Partner Does"
          ]
        },
        {
          "process": "ref/mkt/onboarding-email-program",
          "activities": [
            "Read the Account",
            "Set the Goal for This Run",
            "Brief the Agents that Act",
            "Watch What the Customer Does",
            "Confirm the Customer Is Active"
          ]
        },
        {
          "process": "ref/mkt/pricing-change",
          "activities": [
            "Answer What Comes Back",
            "Watch the Transition"
          ]
        },
        {
          "process": "ref/mkt/product-positioning",
          "activities": [
            "Read the Product's Own Record"
          ]
        },
        {
          "process": "ref/mkt/win-back-campaign",
          "activities": [
            "Read the Named Accounts"
          ]
        },
        {
          "process": "ref/mkt/win-loss-analysis",
          "activities": [
            "Brief the Study",
            "Check Who May Be Approached"
          ]
        }
      ]
    }
  }
}

README.mdWhat the package is, and why it carries no mcp.json.

# account-monitor

This is an abstract agent, published as an Agent Plugin. An abstract
agent is a job description rather than an implementation. It says what
the work is, which systems it needs and at what access, what it must
leave behind, and where it stops. Any agent that leaves the records can
implement it.

The plugin carries seven skills, no `mcp.json`, and everything the
abstract agent owes under `com.agentcatalog.agent/`.

## Why there is no mcp.json

A conformant server entry in `mcp.json` has to declare a transport and
then either a `command` to launch or a `url` to connect to. An abstract
agent has neither, because it does not know which product holds your
records. Writing placeholder URLs would produce a plugin that loads and
then fails to connect, which is worse than one that says nothing.

Section 6.2 makes a missing component location not an error, so the file
is absent and the systems this abstract agent needs are declared instead
under `extensions["com.agentcatalog.agent"].servers`, as a role, an
access level and what the access is for. Whoever builds an agent from
this plugin writes the `mcp.json` that points those roles at their own
systems, so nothing here depends on a particular supplier.

## Why the obligations are in an extension

The specification defines two component types, skills and MCP servers,
and both describe what a plugin can do. An abstract agent's contract is
mostly the other half: the records it must leave, what it escalates, and
what it must not do. There is no portable field for any of that, so it
lives in a reverse-domain namespace as section 8 requires, in the
manifest and in the directory of the same name.

## Where this abstract agent is used

`ref/mkt/account-based-play`, `ref/mkt/advisory-board`,
`ref/mkt/award-entry`, `ref/mkt/churn-risk-play`,
`ref/mkt/customer-case-study`, `ref/mkt/customer-newsletter`, and 9
more. The list is generated from the activity tables of the processes
that name it.

https://agentcatalog.com/abstract-agents/account-monitor

skills/bring-the-account-picture-to-a-briefing/SKILL.mdThe agent sits in the briefing that opens a run and gives every agent that will act the same picture of the account at the same time. Use it when running `ref/mkt/onboarding-email-program` and `ref/mkt/win-loss-analysis`.

---
name: bring-the-account-picture-to-a-briefing
description: The agent sits in the briefing that opens a run and gives every agent that will act the same picture of the account at the same time. Use it when running `ref/mkt/onboarding-email-program` and `ref/mkt/win-loss-analysis`.
license: CC-BY-4.0
metadata:
  agent: account-monitor
  agent-version: "1"
---

# Bring the account picture to a briefing

## What it does

The agent sits in the briefing that opens a run and gives every agent
that will act the same picture of the account at the same time.

## Where it happens

The agent does this in two activities across two reference processes.
Each one names the activity as that process words it.

- **Brief the Agents that Act**
  - `ref/mkt/onboarding-email-program`, activity 5 -
    [Run Onboarding Email Program](../../../../processes/marketing/onboarding-email-program.md)
- **Brief the Study**
  - `ref/mkt/win-loss-analysis`, activity 3 -
    [Conduct Win/Loss Analysis](../../../../processes/marketing/win-loss-analysis.md)

## What to record

For each question it answers, it records the account, what was asked,
the answer, the date, and every system it read with the value it found
and when that value was current. A milestone answer names the milestone
list at its version and says which of those milestones the account has
met. A risk signal is recorded whichever way it was judged, with the
reading that settled it. Anything it could not read is recorded as
unread instead of being left out of the answer.

That contract covers every activity this abstract agent takes on, and it
is repeated in `com.agentcatalog.agent/RECORDS.md`. What the abstract
agent does not do is in `com.agentcatalog.agent/NOT.md`.

skills/check-a-claim-against-the-systems/SKILL.mdThe agent takes a figure or a milestone somebody has stated and reads it back out of the systems that hold the record, saying whether it holds and naming anything it could not check. Use it when running `ref/mkt/customer-case-study` and `ref/mkt/onboarding-email-program`.

---
name: check-a-claim-against-the-systems
description: The agent takes a figure or a milestone somebody has stated and reads it back out of the systems that hold the record, saying whether it holds and naming anything it could not check. Use it when running `ref/mkt/customer-case-study` and `ref/mkt/onboarding-email-program`.
license: CC-BY-4.0
metadata:
  agent: account-monitor
  agent-version: "1"
---

# Check a claim against the systems

## What it does

The agent takes a figure or a milestone somebody has stated and reads it
back out of the systems that hold the record, saying whether it holds
and naming anything it could not check.

## Where it happens

The agent does this in two activities across two reference processes.
Each one names the activity as that process words it.

- **Confirm the Customer Is Active**
  - `ref/mkt/onboarding-email-program`, activity 13 -
    [Run Onboarding Email Program](../../../../processes/marketing/onboarding-email-program.md)
- **Verify the Customer's Results**
  - `ref/mkt/customer-case-study`, activity 7 -
    [Produce Customer Case Study](../../../../processes/marketing/customer-case-study.md)

## What to record

For each question it answers, it records the account, what was asked,
the answer, the date, and every system it read with the value it found
and when that value was current. A milestone answer names the milestone
list at its version and says which of those milestones the account has
met. A risk signal is recorded whichever way it was judged, with the
reading that settled it. Anything it could not read is recorded as
unread instead of being left out of the answer.

That contract covers every activity this abstract agent takes on, and it
is repeated in `com.agentcatalog.agent/RECORDS.md`. What the abstract
agent does not do is in `com.agentcatalog.agent/NOT.md`.

skills/read-an-accounts-standing/SKILL.mdThe agent reads one named account across the systems that hold it, covering product usage, support history, contract terms and where the account sits in the renewal cycle. Use it when running `ref/mkt/account-based-play` and 7 other reference processes.

---
name: read-an-accounts-standing
description: The agent reads one named account across the systems that hold it, covering product usage, support history, contract terms and where the account sits in the renewal cycle. Use it when running `ref/mkt/account-based-play` and 7 other reference processes.
license: CC-BY-4.0
metadata:
  agent: account-monitor
  agent-version: "1"
---

# Read an accounts standing

## What it does

The agent reads one named account across the systems that hold it,
covering product usage, support history, contract terms and where the
account sits in the renewal cycle. Every answer comes back with what
each system said and when that reading was current.

## Where it happens

The agent does this in seven activities across eight reference
processes. Each one names the activity as that process words it.

- **Check Each Account's Standing**
  - `ref/mkt/customer-reference`, activity 4 -
    [Recruit Customer Reference](../../../../processes/marketing/customer-reference.md)
- **Check What the Partner Holds**
  - `ref/mkt/enable-channel-partner`, activity 2 -
    [Enable Channel Partner](../../../../processes/marketing/enable-channel-partner.md)
- **Check Where the Account Stands**
  - `ref/mkt/customer-case-study`, activity 2 -
    [Produce Customer Case Study](../../../../processes/marketing/customer-case-study.md)
  - `ref/mkt/customer-newsletter`, activity 3 -
    [Publish Customer Newsletter](../../../../processes/marketing/customer-newsletter.md)
- **Confirm the Signal**
  - `ref/mkt/churn-risk-play`, activity 2 -
    [Run Churn-Risk Play](../../../../processes/marketing/churn-risk-play.md)
- **Read the Account**
  - `ref/mkt/onboarding-email-program`, activity 2 -
    [Run Onboarding Email Program](../../../../processes/marketing/onboarding-email-program.md)
- **Read the Accounts**
  - `ref/mkt/account-based-play`, activity 2 -
    [Run Account-Based Play](../../../../processes/marketing/account-based-play.md)
- **Read the Product's Own Record**
  - `ref/mkt/product-positioning`, activity 4 -
    [Develop Product Positioning](../../../../processes/marketing/product-positioning.md)

## What to record

For each question it answers, it records the account, what was asked,
the answer, the date, and every system it read with the value it found
and when that value was current. A milestone answer names the milestone
list at its version and says which of those milestones the account has
met. A risk signal is recorded whichever way it was judged, with the
reading that settled it. Anything it could not read is recorded as
unread instead of being left out of the answer.

That contract covers every activity this abstract agent takes on, and it
is repeated in `com.agentcatalog.agent/RECORDS.md`. What the abstract
agent does not do is in `com.agentcatalog.agent/NOT.md`.

skills/record-what-came-back/SKILL.mdThe agent takes the questions, objections and cancellation requests that arrive, reads the account behind each one, and records what was offered, what the account did, and the date. Use it when running `ref/mkt/churn-risk-play` and `ref/mkt/pricing-change`.

---
name: record-what-came-back
description: The agent takes the questions, objections and cancellation requests that arrive, reads the account behind each one, and records what was offered, what the account did, and the date. Use it when running `ref/mkt/churn-risk-play` and `ref/mkt/pricing-change`.
license: CC-BY-4.0
metadata:
  agent: account-monitor
  agent-version: "1"
---

# Record what came back

## What it does

The agent takes the questions, objections and cancellation requests that
arrive, reads the account behind each one, and records what was offered,
what the account did, and the date.

## Where it happens

The agent does this in two activities across two reference processes.
Each one names the activity as that process words it.

- **Answer What Comes Back**
  - `ref/mkt/pricing-change`, activity 11 -
    [Launch Pricing Change](../../../../processes/marketing/pricing-change.md)
- **Record the Outcome**
  - `ref/mkt/churn-risk-play`, activity 11 -
    [Run Churn-Risk Play](../../../../processes/marketing/churn-risk-play.md)

## What to record

For each question it answers, it records the account, what was asked,
the answer, the date, and every system it read with the value it found
and when that value was current. A milestone answer names the milestone
list at its version and says which of those milestones the account has
met. A risk signal is recorded whichever way it was judged, with the
reading that settled it. Anything it could not read is recorded as
unread instead of being left out of the answer.

That contract covers every activity this abstract agent takes on, and it
is repeated in `com.agentcatalog.agent/RECORDS.md`. What the abstract
agent does not do is in `com.agentcatalog.agent/NOT.md`.

skills/say-who-may-be-approached/SKILL.mdThe agent applies the stated criteria, the consent record and each account's standing to say which accounts may be approached, which have to be kept out and why, and which have agreed to be named. Use it when running `ref/mkt/advisory-board` and 5 other reference processes.

---
name: say-who-may-be-approached
description: The agent applies the stated criteria, the consent record and each account's standing to say which accounts may be approached, which have to be kept out and why, and which have agreed to be named. Use it when running `ref/mkt/advisory-board` and 5 other reference processes.
license: CC-BY-4.0
metadata:
  agent: account-monitor
  agent-version: "1"
---

# Say who may be approached

## What it does

The agent applies the stated criteria, the consent record and each
account's standing to say which accounts may be approached, which have
to be kept out and why, and which have agreed to be named.

## Where it happens

The agent does this in six activities across six reference processes.
Each one names the activity as that process words it.

- **Build the Suppression List**
  - `ref/mkt/display-retargeting`, activity 5 -
    [Run Display Retargeting Campaign](../../../../processes/marketing/display-retargeting.md)
- **Check Who May Be Approached**
  - `ref/mkt/win-loss-analysis`, activity 6 -
    [Conduct Win/Loss Analysis](../../../../processes/marketing/win-loss-analysis.md)
- **Choose the Customers to Ask**
  - `ref/mkt/customer-reviews`, activity 4 -
    [Solicit Customer Reviews](../../../../processes/marketing/customer-reviews.md)
- **Get the Customer's Agreement**
  - `ref/mkt/award-entry`, activity 7 -
    [Submit Award Entry](../../../../processes/marketing/award-entry.md)
- **Read the Named Accounts**
  - `ref/mkt/win-back-campaign`, activity 2 -
    [Run Win-Back Campaign](../../../../processes/marketing/win-back-campaign.md)
- **Settle the Membership**
  - `ref/mkt/advisory-board`, activity 3 -
    [Convene Customer Advisory Board](../../../../processes/marketing/advisory-board.md)

## What to record

For each question it answers, it records the account, what was asked,
the answer, the date, and every system it read with the value it found
and when that value was current. A milestone answer names the milestone
list at its version and says which of those milestones the account has
met. A risk signal is recorded whichever way it was judged, with the
reading that settled it. Anything it could not read is recorded as
unread instead of being left out of the answer.

That contract covers every activity this abstract agent takes on, and it
is repeated in `com.agentcatalog.agent/RECORDS.md`. What the abstract
agent does not do is in `com.agentcatalog.agent/NOT.md`.

skills/set-the-milestones-for-a-run/SKILL.mdThe agent writes down the milestones a customer has to reach in a run and says how each one will be read from the systems. Use it when running `ref/mkt/onboarding-email-program`.

---
name: set-the-milestones-for-a-run
description: The agent writes down the milestones a customer has to reach in a run and says how each one will be read from the systems. Use it when running `ref/mkt/onboarding-email-program`.
license: CC-BY-4.0
metadata:
  agent: account-monitor
  agent-version: "1"
---

# Set the milestones for a run

## What it does

The agent writes down the milestones a customer has to reach in a run
and says how each one will be read from the systems.

## Where it happens

The agent does this in one activity across one reference process. Each
one names the activity as that process words it.

- **Set the Goal for This Run**
  - `ref/mkt/onboarding-email-program`, activity 3 -
    [Run Onboarding Email Program](../../../../processes/marketing/onboarding-email-program.md)

## What to record

For each question it answers, it records the account, what was asked,
the answer, the date, and every system it read with the value it found
and when that value was current. A milestone answer names the milestone
list at its version and says which of those milestones the account has
met. A risk signal is recorded whichever way it was judged, with the
reading that settled it. Anything it could not read is recorded as
unread instead of being left out of the answer.

That contract covers every activity this abstract agent takes on, and it
is repeated in `com.agentcatalog.agent/RECORDS.md`. What the abstract
agent does not do is in `com.agentcatalog.agent/NOT.md`.

skills/watch-an-account-after-contact/SKILL.md"Once something has been sent or a change has landed, the agent keeps reading the account and reports what it did: the usage, the replies, the support requests, the deals registered, the milestones reached and where the renewal now stands. Use it when running `ref/mkt/account-based-play` and 4 other reference processes."

---
name: watch-an-account-after-contact
description: "Once something has been sent or a change has landed, the agent keeps reading the account and reports what it did: the usage, the replies, the support requests, the deals registered, the milestones reached and where the renewal now stands. Use it when running `ref/mkt/account-based-play` and 4 other reference processes."
license: CC-BY-4.0
metadata:
  agent: account-monitor
  agent-version: "1"
---

# Watch an account after contact

## What it does

Once something has been sent or a change has landed, the agent keeps
reading the account and reports what it did: the usage, the replies, the
support requests, the deals registered, the milestones reached and where
the renewal now stands.

## Where it happens

The agent does this in five activities across five reference processes.
Each one names the activity as that process words it.

- **Watch Engagement by Account**
  - `ref/mkt/account-based-play`, activity 16 -
    [Run Account-Based Play](../../../../processes/marketing/account-based-play.md)
- **Watch What the Account Does**
  - `ref/mkt/churn-risk-play`, activity 10 -
    [Run Churn-Risk Play](../../../../processes/marketing/churn-risk-play.md)
- **Watch What the Customer Does**
  - `ref/mkt/onboarding-email-program`, activity 10 -
    [Run Onboarding Email Program](../../../../processes/marketing/onboarding-email-program.md)
- **Watch What the Partner Does**
  - `ref/mkt/enable-channel-partner`, activity 14 -
    [Enable Channel Partner](../../../../processes/marketing/enable-channel-partner.md)
- **Watch the Transition**
  - `ref/mkt/pricing-change`, activity 14 -
    [Launch Pricing Change](../../../../processes/marketing/pricing-change.md)

## What to record

For each question it answers, it records the account, what was asked,
the answer, the date, and every system it read with the value it found
and when that value was current. A milestone answer names the milestone
list at its version and says which of those milestones the account has
met. A risk signal is recorded whichever way it was judged, with the
reading that settled it. Anything it could not read is recorded as
unread instead of being left out of the answer.

That contract covers every activity this abstract agent takes on, and it
is repeated in `com.agentcatalog.agent/RECORDS.md`. What the abstract
agent does not do is in `com.agentcatalog.agent/NOT.md`.

mcp.jsonNot in the package.

Every server entry in an mcp.json needs either a command to run or a url to connect to, and the package cannot know either of those until somebody points it at their own systems. So the file is left out, the systems the agent needs are named by role in plugin.json, and you write the mcp.json when you set it up.

Setting it up

Point it at your systems

To run an agent built from this package, write an mcp.json that says which of your systems fills each role. The package names these 5 roles, and the access level beside each one is the most the work needs.

product-usage-data · readsupport-system · readcrm · readcontract-and-renewal-records · readaccount-standing-record · write

mcp.json You write this file. It is not part of the package.

{
  "$schema": "https://agent-plugins.org/schemas/1.1.0/mcp.schema.json",
  "mcpServers": {
    "product-usage-data": {
      "type": "streamable-http",
      "url": "https://mcp.yourcompany.com/product-usage-data"
    },
    "support-system": {
      "type": "streamable-http",
      "url": "https://mcp.yourcompany.com/support-system"
    },
    "crm": {
      "type": "streamable-http",
      "url": "https://mcp.yourcompany.com/crm"
    },
    "contract-and-renewal-records": {
      "type": "streamable-http",
      "url": "https://mcp.yourcompany.com/contract-and-renewal-records"
    },
    "account-standing-record": {
      "type": "streamable-http",
      "url": "https://mcp.yourcompany.com/account-standing-record"
    }
  }
}

The hostnames here are a placeholder for your own. Whichever product holds each of these systems, the agent you build is yours and nobody else has to be involved in running it.

Where it works

36 processes, 68 activities

The highlighted activities are the ones this abstract agent does. The other activities in each process belong to other abstract agents or to people.

Run Account-Based Playref/mkt/account-based-play · 19 activities

Read the Accounts · Watch Engagement by Account

Convene Customer Advisory Boardref/mkt/advisory-board · 16 activities

Settle the Membership

Submit Award Entryref/mkt/award-entry · 16 activities

Get the Customer's Agreement

Run Churn-Risk Playref/mkt/churn-risk-play · 12 activities

Confirm the Signal · Watch What the Account Does · Record the Outcome

Produce Customer Case Studyref/mkt/customer-case-study · 21 activities

Check Where the Account Stands · Verify the Customer's Results

Publish Customer Newsletterref/mkt/customer-newsletter · 17 activities

Check Where the Account Stands

Recruit Customer Referenceref/mkt/customer-reference · 15 activities

Check Each Account's Standing

Solicit Customer Reviewsref/mkt/customer-reviews · 16 activities

Choose the Customers to Ask

Run Display Retargeting Campaignref/mkt/display-retargeting · 16 activities

Build the Suppression List

Enable Channel Partnerref/mkt/enable-channel-partner · 16 activities

Check What the Partner Holds · Watch What the Partner Does

Run Onboarding Email Programref/mkt/onboarding-email-program · 14 activities

Read the Account · Set the Goal for This Run · Brief the Agents that Act · Watch What the Customer Does · Confirm the Customer Is Active

Launch Pricing Changeref/mkt/pricing-change · 16 activities

Answer What Comes Back · Watch the Transition

Develop Product Positioningref/mkt/product-positioning · 17 activities

Read the Product's Own Record

Run Win-Back Campaignref/mkt/win-back-campaign · 20 activities

Read the Named Accounts

Conduct Win/Loss Analysisref/mkt/win-loss-analysis · 21 activities

Brief the Study · Check Who May Be Approached

Match It to One Customer · Check the Direct Pipeline · Watch the Registered Deal

Resolve a Channel Conflictref/prt/resolve-a-channel-conflict · 15 activities

Match the Two Records · Pull What Each Side Holds

Run a Co-Sellref/prt/run-a-co-sell · 17 activities

Read the Account on Both Sides · Record the Outcome against the Partner

Model the Territoriesref/rev/model-the-territories · 15 activities

Size Each Account

Move an Accountref/rev/move-an-account · 14 activities

Check the Account's Standing · Read the Work Already Done · Hand the Account Over · Watch the First Weeks

Publish the Carveref/rev/publish-the-carve · 15 activities

Test Each Appeal Against the Record

Answer an Inbound Inquiryref/sls/answer-an-inbound-inquiry · 14 activities

Read the Account

Book the Meetingref/sls/book-the-meeting · 16 activities

Check the Account and the Owner

Book the Orderref/sls/book-the-order · 16 activities

Check Credit and Standing

Build a Prospect Listref/sls/build-a-prospect-list · 14 activities

Check Who Already Owns Them · Read Each Account · Set the Refresh

Close a Renewalref/sls/close-a-renewal · 15 activities

Tell the People Who Run the Account

Hand over to Deliveryref/sls/hand-over-to-delivery · 16 activities

Read Where the Account Stands

Map the People at an Accountref/sls/map-the-account-people · 14 activities

Watch for People Moving

Open a Renewalref/sls/open-a-renewal · 13 activities

Read the Account's Standing · Score the Risk of Leaving

Qualify an Opportunityref/sls/qualify-an-opportunity · 16 activities

Read the Account Again

Research an Accountref/sls/research-an-account · 15 activities

Read What We Already Hold · Map What the Account Already Owns · Read the History Between Us · Check the Account May Be Worked · Set the Refresh Date

Return a Lead to Nurtureref/sls/return-a-lead-to-nurture · 13 activities

Name What Would Bring It Back Early · Set the Watch · Look Again on the Date

Review the Account Plansref/sls/review-the-account-plans · 14 activities

Read What Actually Happened · Check Whether the People Changed

Work a Churn Riskref/sls/work-a-churn-risk · 15 activities

Confirm the Signal · Read the Account's Standing · Do the Work

Work a Referralref/sls/work-a-referral · 14 activities

Check the Account · Record the Referral

Write an Account Planref/sls/write-an-account-plan · 16 activities

Read the Account Read and the Map · Set the Review Date and the Tripwires

The records it leaves

What it writes down

For each question it answers, it records the account, what was asked, the answer, the date, and every system it read with the value it found and when that value was current. A milestone answer names the milestone list at its version and says which of those milestones the account has met. A risk signal is recorded whichever way it was judged, with the reading that settled it. Anything it could not read is recorded as unread instead of being left out of the answer.

Where it stops

What it will not do

Does not contact the customer and does not send anything to them. Does not decide what to do about what it finds; it answers the question and the agents that act read the answer. Does not choose which messages a customer receives or run a standing program for them, which is the lifecycle-manager agent. Does not ask a customer to do anything in public, which is the advocacy-manager agent, and that agent asks this one whether the account is in standing before it makes the ask. Does not score or route an inbound lead, which is the lead-scorer agent. Does not measure a campaign or a published piece, which is the analytics agent, and does not own the department's report for a period, which is the reporting-manager agent; this agent answers about one named account. Does not change the CRM, the contract, or anything else it reads. Does not negotiate a renewal or make a commercial offer, which stays with the people who own the account.

These limits are written into the job on purpose, so that everyone adopting the process knows which decisions stay with a person.

Agents built from this plugin

Who has built one, and how far each one got

Nobody has built one yet. The catalog says so rather than leaving the section off, because an empty list is a fact about the document.

declare it implements: account-monitor v1 One line in your agent's manifest. The catalog checks it against the records your runs actually leave.

A guess, not a claim

The closest things in the catalog, by description

Nobody above has claimed any of these. This is the nearest match by wording between the job written above and what publishers wrote about their own listings, closest first, and wording is not evidence: some of these will be wrong, and a job with nothing suitable still gets whatever was nearest. The publishers have never heard of this abstract agent and have promised nothing about it. What a claim looks like is in the table above, and it is signed.

Might do the job

Things that could be the worker: an agent, a packaged product, a plugin.

The Website Specification — A2A agent Agent · specification.website Read-only A2A agent that answers natural-language questions about The Website Specification and returns matching topics with status, canonical URL,…

Might be reached for

Not workers. These are capability that whatever does the job would use, which is a different relationship and kept in its own column for that reason.

Phano Intelligence Platform MCP server · phano.ai · hosted · mirrored from the MCP registry Read-only account intelligence for Customer Success and Account Management teams.
Agentic Diaries Audit MCP server · kandikandikandi.github.io · hosted · mirrored from the MCP registry Checks what a support agent knew but did not say against its policies. Free tier, no login.
fundzwatch MCP server · fund-z.github.io · run it yourself · mirrored from the MCP registry Business-event intelligence for AI agents: funding, exec moves, UCC and benefit-plan signals.