Use cases/An agent buys context it can trust

An agent buys context it can trust

ProposedNo plan row yetContextFileAI Source of TruthPromptCompute

Not in the plan. ContextFile's quarantine and AI Source of Truth's verdicts are merged.

Use case

An agent buys context it can trust: a workflow across ContextFile, AI Source of Truth, PromptCompute that starts from an agent asks for context it does not have.

Who would use it

Builders whose agents need outside docs, datasets or skills.

What value it provides

The agent buys a ContextFile artifact under a spending limit. The AI Source of Truth verdict shows before purchase, the artifact sits in quarantine until it passes a check, and larger purchases wait for a person.

Why it matters to the user

Your agents can use outside knowledge without you reviewing every file, and you can see what each one cost and whether it helped.

The problem

Agents pull context from anywhere. A bad source poisons the work, and nothing guards the payment.

Expected outcomes

Purchases held in quarantine that failed their check
Spend per artifact against its outcome

Each run records its outcome as met, not met or not measured.

How it runs

1. Find candidate artifactsContextFileBuilt

2. Show the verdict on each version's recordAI Source of TruthBuilt

3. Hold purchases above the limit for a personPromptComputeBuilt

4. Quarantine until it passes a checkContextFileBuilt

5. Report how the purchase turned outContextFileBuilt

Products and connections

Product	Role in this workflow	Docs
ContextFile	Find candidate artifacts; Quarantine until it passes a check; Report how the purchase turned out	Reference · ContextFile docs
AI Source of Truth	Show the verdict on each version's record	Reference · AI Source of Truth docs
PromptCompute	Hold purchases above the limit for a person	Reference · PromptCompute docs

Outside the family: A payment method on ContextFile.

What a person approves

Action	Default risk
`agent.order.place`	high
`escrow.release`	high

Action keys are from the PromptCompute catalog.

How to access it from an agent

An agent will run this workflow through the PromptReports MCP server or API, with an ID agent token. The runner will then call each product for it. Proposed

// MCP: https://mcp.promptreports.ai/mcp (Proposed)
start_run({
  workflow: "buy-context-you-can-trust",
  inputs: { "need": "Stripe Billing docs, 2026 version", "limit": "$50 per purchase" },
  dry_run: false
})
// returns { run_id, status: "running", approvals_url }
get_run({ run_id })   // poll, or subscribe to run events

curl -X POST https://api.promptreports.ai/v1/runs \
  -H "Authorization: Bearer $PROMPTREPORTS_AGENT_TOKEN" \
  -d '{"workflow":"buy-context-you-can-trust"}'

Or call the products yourself, step by step:

Product	Endpoint	Tools
ContextFile	Built	to confirm
AI Source of Truth	`https://aisourceoftruth.com/v1/family/facts`	to confirm
PromptCompute	`https://mcp.promptcompute.ai/mcp`	`request_approval` `get_approval` `cancel_approval` `answer_approver`

Set it up

Sign in at app.promptreports.ai with PromptReports ID, and pick your org.
Connect ContextFile, AI Source of Truth, PromptCompute on the Connections page. Link A payment method on ContextFile through the product that uses it.
Register the agent that will run it on the Agents page, and give it a version manifest.
Set the approval policy for agent.order.place, escrow.release. Start with Require approval; critical actions are never auto.
Fill in the inputs (Need, Limit) and choose the trigger: an agent asks for context it does not have.
Start a dry run. It will count no workflow request, and it shows which steps would record missing evidence.
Start the run. Approve each gate in the approvals inbox, then read the signed record when it ends.