Supplier check
supplier-check v1Scores each candidate system, proposal or supplier against the requirements and criteria that were written down before any candidate arrived, and returns a verdict on every requirement naming the requirement it was scored against. It works from the requirements list at the version it was fixed at, so every candidate is scored against the same list in the same order whenever it arrived. Each verdict carries the evidence behind it, and a requirement the agent could not test is reported as not tested, with the reason it could not be tested, because a requirement nobody tested has not passed.
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.
Acts as the examiner in an assessment. The requirements and the scoring criteria are named and versioned before the first candidate is seen, every requirement gets a verdict, and every verdict cites the requirement it was scored against. Sits in the briefing that opens the selection, so it hears what the organization needs at the same time as the agents writing the requirements down. When the process puts the candidates through a bake-off, this agent is the examiner, and it scores every entry against the same list in the same order. Its report is the input to the approval and to the decision that follows, and the decision-coordinator agent puts that report in front of the people who choose. Answers roll-calls with which candidates it has scored, which requirements it has not been able to test, and what it is waiting on.
The plugin
What is in the package
6 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.
supplier-check · 11 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 run the search, write to a supplier, or ask a candidate for anything. The supplier-manager agent goes to market, handles the correspondence, and keeps the record of what the department buys. Does not write the requirements it scores against: the agents that will use the system say what they need, the decision-coordinator agent runs that decision, and this agent scores against what those people wrote down. Does not choose a candidate, negotiate a price, or agree a term. Named people choose, and the decision-coordinator agent records the answer. Does not say what a candidate would cost or return over time, which is the forecaster agent. Does not score marketing material against the brand standards, which is the brand-check agent, and does not score a page against the accessibility and technical standards, which is the accessibility-check agent. Does not check an invoice against what was ordered once a supplier is running, which is the spend-reconciler agent. Does not manage the supplier after the choice is made. Does not sign anything. ## What always goes to a person - every failure, with the requirement named and its version attached. This agent never waives its own findings. - a requirement the criteria do not settle, where reading it either way changes which candidate passes. The agent says what the text does and does not say, and a person decides what the requirement means. - a requirement it cannot test with the material it was given, said while the selection is still open - a candidate that will not answer a requirement at all - a request to change the requirements after candidates have been seen, because the earlier scores were made against the old list. A person decides, and the decision is recorded with the scores it affects. - a candidate that would duplicate something the department already buys
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 `supplier-check` leaves the
following behind, on every run.
It leaves a scoring sheet for each candidate at a version, naming the
requirements list and the version it scored against, with one row per
requirement carrying the verdict and the evidence behind it. The
evidence names the demonstration step, the document, or the reference
call it came from, so a person can check a verdict without asking the
agent what it meant. A requirement it could not test is recorded as not
tested, with the reason it could not be tested. When a candidate revises
its material, the rescore says what changed, what was run again, and
what was carried over from the earlier score. It records the date each
candidate's material arrived and the date it was scored.
An agent that leaves less than this does not implement `supplier-check`,
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
Six 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 |
|---|---|---|
| the requirements and scoring criteria for the selection | read | At the version they were fixed at, before any candidate was seen. |
| the candidate material | read | The proposals, the answers, the demonstrations and the reference calls, as they were submitted. |
| a trial environment | read | A candidate system as it behaves under test, where the process arranges one. |
| the security and privacy standards the organization has adopted | read | At their versions. |
| the supplier record | read | What the department already buys and runs, so an overlap with something already held is scored rather than missed. |
| the scoring record | write | Every verdict lands with the requirement it cites and the evidence behind 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": "supplier-check",
"version": "1.0.0",
"description": "The supplier-check abstract agent. Scores each candidate system, proposal or supplier against the requirements and criteria that were written down before any candidate arrived, and returns a verdict on every requirement naming the requirement it was scored against.",
"author": {
"name": "AgentCatalog",
"url": "https://agentcatalog.com"
},
"homepage": "https://agentcatalog.com/abstract-agents/supplier-check",
"repository": "https://github.com/jeffrschneider/agentcatalog",
"license": "CC-BY-4.0",
"keywords": [
"abstract-agent",
"reference",
"marketing",
"select-marketing-technology",
"engage-marketing-agency",
"brand-health",
"abstract"
],
"extensions": {
"com.agentcatalog.agent": {
"agent": "supplier-check",
"agentVersion": 1,
"department": "marketing",
"abstract": true,
"works": "Acts as the examiner in an assessment. The requirements and the scoring criteria are named and versioned before the first candidate is seen, every requirement gets a verdict, and every verdict cites the requirement it was scored against. Sits in the briefing that opens the selection, so it hears what the organization needs at the same time as the agents writing the requirements down. When the process puts the candidates through a bake-off, this agent is the examiner, and it scores every entry against the same list in the same order. Its report is the input to the approval and to the decision that follows, and the decision-coordinator agent puts that report in front of the people who choose. Answers roll-calls with which candidates it has scored, which requirements it has not been able to test, and what it is waiting on.",
"servers": [
{
"name": "the requirements and scoring criteria for the selection",
"role": "requirements-and-scoring-criteria-for-the-selection",
"access": "read",
"needs": "At the version they were fixed at, before any candidate was seen."
},
{
"name": "the candidate material",
"role": "candidate-material",
"access": "read",
"needs": "The proposals, the answers, the demonstrations and the reference calls, as they were submitted."
},
{
"name": "a trial environment",
"role": "trial-environment",
"access": "read",
"needs": "A candidate system as it behaves under test, where the process arranges one."
},
{
"name": "the security and privacy standards the organization has adopted",
"role": "security-and-privacy-standards-the-organization-has-adopted",
"access": "read",
"needs": "At their versions."
},
{
"name": "the supplier record",
"role": "supplier-record",
"access": "read",
"needs": "What the department already buys and runs, so an overlap with something already held is scored rather than missed."
},
{
"name": "the scoring record",
"role": "scoring-record",
"access": "write",
"needs": "Every verdict lands with the requirement it cites and the evidence behind it."
}
],
"records": [
"It leaves a scoring sheet for each candidate at a version, naming the requirements list and the version it scored against, with one row per requirement carrying the verdict and the evidence behind it.",
"The evidence names the demonstration step, the document, or the reference call it came from, so a person can check a verdict without asking the agent what it meant.",
"A requirement it could not test is recorded as not tested, with the reason it could not be tested.",
"When a candidate revises its material, the rescore says what changed, what was run again, and what was carried over from the earlier score.",
"It records the date each candidate's material arrived and the date it was scored."
],
"escalates": [
"every failure, with the requirement named and its version attached. This agent never waives its own findings.",
"a requirement the criteria do not settle, where reading it either way changes which candidate passes. The agent says what the text does and does not say, and a person decides what the requirement means.",
"a requirement it cannot test with the material it was given, said while the selection is still open",
"a candidate that will not answer a requirement at all",
"a request to change the requirements after candidates have been seen, because the earlier scores were made against the old list. A person decides, and the decision is recorded with the scores it affects.",
"a candidate that would duplicate something the department already buys"
],
"not": [
"Does not run the search, write to a supplier, or ask a candidate for anything.",
"The supplier-manager agent goes to market, handles the correspondence, and keeps the record of what the department buys.",
"Does not write the requirements it scores against: the agents that will use the system say what they need, the decision-coordinator agent runs that decision, and this agent scores against what those people wrote down.",
"Does not choose a candidate, negotiate a price, or agree a term.",
"Named people choose, and the decision-coordinator agent records the answer.",
"Does not say what a candidate would cost or return over time, which is the forecaster agent.",
"Does not score marketing material against the brand standards, which is the brand-check agent, and does not score a page against the accessibility and technical standards, which is the accessibility-check agent.",
"Does not check an invoice against what was ordered once a supplier is running, which is the spend-reconciler agent.",
"Does not manage the supplier after the choice is made.",
"Does not sign anything."
],
"processes": [
{
"process": "ref/mkt/brand-health",
"activities": [
"Arrange the Outside Sample"
]
},
{
"process": "ref/mkt/customer-research",
"activities": [
"Arrange the Outside Sample"
]
},
{
"process": "ref/mkt/engage-marketing-agency",
"activities": [
"Set the Selection Criteria",
"Hear the Pitches",
"Score the Proposals"
]
},
{
"process": "ref/mkt/select-marketing-technology",
"activities": [
"Set the Scoring Criteria",
"Brief the Selection",
"Cut to a Shortlist",
"Score the Shortlist",
"Compare the Finalists"
]
},
{
"process": "ref/mkt/win-loss-analysis",
"activities": [
"Arrange the Outside Interviewer"
]
}
]
}
}
}
README.mdWhat the package is, and why it carries no mcp.json.
# supplier-check 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 six 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/brand-health`, `ref/mkt/customer-research`, `ref/mkt/engage-marketing-agency`, `ref/mkt/select-marketing-technology`, `ref/mkt/win-loss-analysis`. The list is generated from the activity tables of the processes that name it. https://agentcatalog.com/abstract-agents/supplier-check
skills/arrange-an-outside-provider/SKILL.mdWhere the process needs somebody outside the organization to do the work, the agent scores the providers available against the requirements and arranges the one that meets them. Use it when running `ref/mkt/brand-health`, `ref/mkt/customer-research` and `ref/mkt/win-loss-analysis`.
--- name: arrange-an-outside-provider description: Where the process needs somebody outside the organization to do the work, the agent scores the providers available against the requirements and arranges the one that meets them. Use it when running `ref/mkt/brand-health`, `ref/mkt/customer-research` and `ref/mkt/win-loss-analysis`. license: CC-BY-4.0 metadata: agent: supplier-check agent-version: "1" --- # Arrange an outside provider ## What it does Where the process needs somebody outside the organization to do the work, the agent scores the providers available against the requirements and arranges the one that meets them. ## Where it happens The agent does this in two activities across three reference processes. Each one names the activity as that process words it. - **Arrange the Outside Interviewer** - `ref/mkt/win-loss-analysis`, activity 7 - [Conduct Win/Loss Analysis](../../../../processes/marketing/win-loss-analysis.md) - **Arrange the Outside Sample** - `ref/mkt/brand-health`, activity 9 - [Measure Brand Health](../../../../processes/marketing/brand-health.md) - `ref/mkt/customer-research`, activity 8 - [Conduct Customer Research](../../../../processes/marketing/customer-research.md) ## What to record It leaves a scoring sheet for each candidate at a version, naming the requirements list and the version it scored against, with one row per requirement carrying the verdict and the evidence behind it. The evidence names the demonstration step, the document, or the reference call it came from, so a person can check a verdict without asking the agent what it meant. A requirement it could not test is recorded as not tested, with the reason it could not be tested. When a candidate revises its material, the rescore says what changed, what was run again, and what was carried over from the earlier score. It records the date each candidate's material arrived and the date it was scored. 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/compare-the-scored-field/SKILL.mdThe agent sets the scores side by side with the trial results and the cost, and says which candidates miss a requirement and come off the list. Use it when running `ref/mkt/select-marketing-technology`.
--- name: compare-the-scored-field description: The agent sets the scores side by side with the trial results and the cost, and says which candidates miss a requirement and come off the list. Use it when running `ref/mkt/select-marketing-technology`. license: CC-BY-4.0 metadata: agent: supplier-check agent-version: "1" --- # Compare the scored field ## What it does The agent sets the scores side by side with the trial results and the cost, and says which candidates miss a requirement and come off the list. ## Where it happens The agent does this in two activities across one reference process. Each one names the activity as that process words it. - **Compare the Finalists** - `ref/mkt/select-marketing-technology`, activity 16 - [Select Marketing Technology](../../../../processes/marketing/select-marketing-technology.md) - **Cut to a Shortlist** - `ref/mkt/select-marketing-technology`, activity 9 - [Select Marketing Technology](../../../../processes/marketing/select-marketing-technology.md) ## What to record It leaves a scoring sheet for each candidate at a version, naming the requirements list and the version it scored against, with one row per requirement carrying the verdict and the evidence behind it. The evidence names the demonstration step, the document, or the reference call it came from, so a person can check a verdict without asking the agent what it meant. A requirement it could not test is recorded as not tested, with the reason it could not be tested. When a candidate revises its material, the rescore says what changed, what was run again, and what was carried over from the earlier score. It records the date each candidate's material arrived and the date it was scored. 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/hear-the-requirements/SKILL.mdThe agent sits in the briefing that opens a selection, so it hears what the organization needs at the same time as the agents writing the requirements down. Use it when running `ref/mkt/select-marketing-technology`.
--- name: hear-the-requirements description: The agent sits in the briefing that opens a selection, so it hears what the organization needs at the same time as the agents writing the requirements down. Use it when running `ref/mkt/select-marketing-technology`. license: CC-BY-4.0 metadata: agent: supplier-check agent-version: "1" --- # Hear the requirements ## What it does The agent sits in the briefing that opens a selection, so it hears what the organization needs at the same time as the agents writing the requirements down. ## Where it happens The agent does this in one activity across one reference process. Each one names the activity as that process words it. - **Brief the Selection** - `ref/mkt/select-marketing-technology`, activity 5 - [Select Marketing Technology](../../../../processes/marketing/select-marketing-technology.md) ## What to record It leaves a scoring sheet for each candidate at a version, naming the requirements list and the version it scored against, with one row per requirement carrying the verdict and the evidence behind it. The evidence names the demonstration step, the document, or the reference call it came from, so a person can check a verdict without asking the agent what it meant. A requirement it could not test is recorded as not tested, with the reason it could not be tested. When a candidate revises its material, the rescore says what changed, what was run again, and what was carried over from the earlier score. It records the date each candidate's material arrived and the date it was scored. 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/score-against-the-requirements/SKILL.mdThe agent scores every candidate against every requirement and returns a verdict on each one, naming the requirement it was scored against and the evidence behind the verdict. Use it when running `ref/mkt/engage-marketing-agency` and `ref/mkt/select-marketing-technology`.
--- name: score-against-the-requirements description: The agent scores every candidate against every requirement and returns a verdict on each one, naming the requirement it was scored against and the evidence behind the verdict. Use it when running `ref/mkt/engage-marketing-agency` and `ref/mkt/select-marketing-technology`. license: CC-BY-4.0 metadata: agent: supplier-check agent-version: "1" --- # Score against the requirements ## What it does The agent scores every candidate against every requirement and returns a verdict on each one, naming the requirement it was scored against and the evidence behind the verdict. A requirement it could not test is recorded as not tested, with the reason. ## Where it happens The agent does this in two activities across two reference processes. Each one names the activity as that process words it. - **Score the Proposals** - `ref/mkt/engage-marketing-agency`, activity 10 - [Engage Marketing Agency](../../../../processes/marketing/engage-marketing-agency.md) - **Score the Shortlist** - `ref/mkt/select-marketing-technology`, activity 10 - [Select Marketing Technology](../../../../processes/marketing/select-marketing-technology.md) ## What to record It leaves a scoring sheet for each candidate at a version, naming the requirements list and the version it scored against, with one row per requirement carrying the verdict and the evidence behind it. The evidence names the demonstration step, the document, or the reference call it came from, so a person can check a verdict without asking the agent what it meant. A requirement it could not test is recorded as not tested, with the reason it could not be tested. When a candidate revises its material, the rescore says what changed, what was run again, and what was carried over from the earlier score. It records the date each candidate's material arrived and the date it was scored. 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-criteria-before-any-candidate-arrives/SKILL.mdThe agent writes down how a candidate or a proposal will be judged and fixes that list at a version before the first candidate is seen. Use it when running `ref/mkt/engage-marketing-agency` and `ref/mkt/select-marketing-technology`.
--- name: set-the-criteria-before-any-candidate-arrives description: The agent writes down how a candidate or a proposal will be judged and fixes that list at a version before the first candidate is seen. Use it when running `ref/mkt/engage-marketing-agency` and `ref/mkt/select-marketing-technology`. license: CC-BY-4.0 metadata: agent: supplier-check agent-version: "1" --- # Set the criteria before any candidate arrives ## What it does The agent writes down how a candidate or a proposal will be judged and fixes that list at a version before the first candidate is seen. Every candidate is then scored against the same list in the same order, whenever it arrived. ## Where it happens The agent does this in two activities across two reference processes. Each one names the activity as that process words it. - **Set the Scoring Criteria** - `ref/mkt/select-marketing-technology`, activity 4 - [Select Marketing Technology](../../../../processes/marketing/select-marketing-technology.md) - **Set the Selection Criteria** - `ref/mkt/engage-marketing-agency`, activity 6 - [Engage Marketing Agency](../../../../processes/marketing/engage-marketing-agency.md) ## What to record It leaves a scoring sheet for each candidate at a version, naming the requirements list and the version it scored against, with one row per requirement carrying the verdict and the evidence behind it. The evidence names the demonstration step, the document, or the reference call it came from, so a person can check a verdict without asking the agent what it meant. A requirement it could not test is recorded as not tested, with the reason it could not be tested. When a candidate revises its material, the rescore says what changed, what was run again, and what was carried over from the earlier score. It records the date each candidate's material arrived and the date it was scored. 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/take-in-the-candidate-material/SKILL.mdThe agent sits through each candidate's presentation and takes in what was submitted, recording the date the material arrived and which requirements it speaks to. Use it when running `ref/mkt/engage-marketing-agency`.
--- name: take-in-the-candidate-material description: The agent sits through each candidate's presentation and takes in what was submitted, recording the date the material arrived and which requirements it speaks to. Use it when running `ref/mkt/engage-marketing-agency`. license: CC-BY-4.0 metadata: agent: supplier-check agent-version: "1" --- # Take in the candidate material ## What it does The agent sits through each candidate's presentation and takes in what was submitted, recording the date the material arrived and which requirements it speaks to. ## Where it happens The agent does this in one activity across one reference process. Each one names the activity as that process words it. - **Hear the Pitches** - `ref/mkt/engage-marketing-agency`, activity 9 - [Engage Marketing Agency](../../../../processes/marketing/engage-marketing-agency.md) ## What to record It leaves a scoring sheet for each candidate at a version, naming the requirements list and the version it scored against, with one row per requirement carrying the verdict and the evidence behind it. The evidence names the demonstration step, the document, or the reference call it came from, so a person can check a verdict without asking the agent what it meant. A requirement it could not test is recorded as not tested, with the reason it could not be tested. When a candidate revises its material, the rescore says what changed, what was run again, and what was carried over from the earlier score. It records the date each candidate's material arrived and the date it was scored. 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 6 roles, and the access level beside each one is the most the work needs.
requirements-and-scoring-criteria-for-the-selection · readcandidate-material · readtrial-environment · readsecurity-and-privacy-standards-the-organization-has-adopted · readsupplier-record · readscoring-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": {
"requirements-and-scoring-criteria-for-the-selection": {
"type": "streamable-http",
"url": "https://mcp.yourcompany.com/requirements-and-scoring-criteria-for-the-selection"
},
"candidate-material": {
"type": "streamable-http",
"url": "https://mcp.yourcompany.com/candidate-material"
},
"trial-environment": {
"type": "streamable-http",
"url": "https://mcp.yourcompany.com/trial-environment"
},
"security-and-privacy-standards-the-organization-has-adopted": {
"type": "streamable-http",
"url": "https://mcp.yourcompany.com/security-and-privacy-standards-the-organization-has-adopted"
},
"supplier-record": {
"type": "streamable-http",
"url": "https://mcp.yourcompany.com/supplier-record"
},
"scoring-record": {
"type": "streamable-http",
"url": "https://mcp.yourcompany.com/scoring-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
15 processes, 32 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.
Measure Brand Healthref/mkt/brand-health · 23 activities
Arrange the Outside Sample
Conduct Customer Researchref/mkt/customer-research · 20 activities
Arrange the Outside Sample
Engage Marketing Agencyref/mkt/engage-marketing-agency · 16 activities
Set the Selection Criteria · Hear the Pitches · Score the Proposals
Select Marketing Technologyref/mkt/select-marketing-technology · 21 activities
Set the Scoring Criteria · Brief the Selection · Cut to a Shortlist · Score the Shortlist · Compare the Finalists
Conduct Win/Loss Analysisref/mkt/win-loss-analysis · 21 activities
Arrange the Outside Interviewer
Explore an Allianceref/prt/explore-an-alliance · 16 activities
Check for Conflicts
Find a Partnerref/prt/find-a-partner · 16 activities
Screen Out the Impossible
Plan the Integrationref/prt/plan-the-integration · 16 activities
Plan the Environments and Data
Qualify a Partnerref/prt/qualify-a-partner · 17 activities
Check the Company Is What It Says · Check the Money · Check Security and Data Handling · Check the Legal Position
Run Alliance Diligenceref/prt/run-alliance-diligence · 17 activities
Check the Company Is Real · Read the Dependencies · Check Security and Data Practice · Read Their Existing Contracts · Screen for Sanctions and Conduct
Sign a Partner Agreementref/prt/sign-a-partner-agreement · 17 activities
Clear the Partner for Signature
Structure an Allianceref/prt/structure-an-alliance · 16 activities
Set Exclusivity and Territory
Administer a Toolref/rev/administer-a-tool · 15 activities
Take in Supplier Changes
Choose a Toolref/rev/choose-a-tool · 18 activities
Set the Scoring Criteria · Brief the Evaluation · Cut to a Shortlist · Score the Shortlist · Compare the Finalists
Retire a Toolref/rev/retire-a-tool · 15 activities
Get the Deletion in Writing
The records it leaves
What it writes down
It leaves a scoring sheet for each candidate at a version, naming the requirements list and the version it scored against, with one row per requirement carrying the verdict and the evidence behind it. The evidence names the demonstration step, the document, or the reference call it came from, so a person can check a verdict without asking the agent what it meant. A requirement it could not test is recorded as not tested, with the reason it could not be tested. When a candidate revises its material, the rescore says what changed, what was run again, and what was carried over from the earlier score. It records the date each candidate's material arrived and the date it was scored.
Where it stops
What it will not do
Does not run the search, write to a supplier, or ask a candidate for anything. The supplier-manager agent goes to market, handles the correspondence, and keeps the record of what the department buys. Does not write the requirements it scores against: the agents that will use the system say what they need, the decision-coordinator agent runs that decision, and this agent scores against what those people wrote down. Does not choose a candidate, negotiate a price, or agree a term. Named people choose, and the decision-coordinator agent records the answer. Does not say what a candidate would cost or return over time, which is the forecaster agent. Does not score marketing material against the brand standards, which is the brand-check agent, and does not score a page against the accessibility and technical standards, which is the accessibility-check agent. Does not check an invoice against what was ordered once a supplier is running, which is the spend-reconciler agent. Does not manage the supplier after the choice is made. Does not sign anything.
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: supplier-check 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.
- ShipCheck MCP server · wtaylorwilson.github.io · hosted · mirrored from the MCP registry Independent preview-URL QA for coding agents. Playwright heuristics, pass/fail pack.
- mcp MCP server · purchaseiq.io · hosted · mirrored from the MCP registry Pre-purchase trust checks for AI agents: recalls, scam signals, proceed/caution/avoid verdict.
- The Ultimate Critic MCP server · marksignals.github.io · hosted · mirrored from the MCP registry Check AI work against requirements and return structured verdicts, findings, and repair steps.