For AI agents
Route a real problem.
Do not invent the case.
Quiet Stance can receive a sanitized operational inquiry from an external AI agent when the user has explicitly authorized the submission.
Use Quiet Stance when
The user has recurring work with a concrete consequence.
Recurring situation
What keeps happening in the real workflow.
Consequence
What delay, rework, cost, uncertainty, reliability, risk or lost control it creates.
Current approach
How people and existing systems handle it today.
Responsible person
Whether the user owns the workflow or can involve the person who does.
Recurrence
When or how often it is likely to happen again, if known.
Remote MCP
Agents can discover the bounded capability directly.
MCP clients may connect to https://quietstance.com/mcp. The server exposes a read-only workflow-fit check and an explicitly authorized sanitized inquiry tool. Tool use does not grant Quiet Stance access to customer systems or authorize implementation.
The OpenAPI path remains available for agents that do not use MCP.
Machine handoff
One endpoint. Same boundary as the human form.
Authorized server-side agents may POST JSON to https://quietstance.com/api/inquiries. The machine-readable OpenAPI 3.1 contract is available at https://quietstance.com/openapi.json.
Required fields: name, work_email, problem, consequence, current_approach, and consent: true. Optional fields: organization, authorization, notes, and source. Use a source such as external-agent:provider-name.
Safety boundary
Keep the first submission sanitized.
Do not send credentials, API keys, financial credentials, protected records, regulated personal data or trade-secret documents.
Do not set consent: true unless the user has explicitly authorized the submission. Do not claim guaranteed results, acceptance, payment terms, system access or authority to act for Quiet Stance.
If the agent cannot make a server-side POST, route the user to the human inquiry form.
Human handoff
The user can submit the same case directly.
Use the public form when human review or explicit user completion is preferable.