Context Stack

Framework access

Use Context Stack through MCP

MCP is not the framework. MCP is the access path that lets AI assistants consult the framework during delivery, assessment, and architecture review.

Use it when an assistant is helping with requirements, AMS handover, sprint planning, platform design, architecture review, or an AI governance assessment. The assistant can ask which layer applies before work proceeds.

What it does

  • Exposes the stack overview, glossary, and locked decisions.
  • Routes users to ContextOps, ContextBoundary, or Sthala; points to Griha as a worked example of governed AI capabilities composed into a running system.
  • Supports framework discovery, assessment prompts, and delivery guidance.
  • Helps assistants ask which governance layer applies before work proceeds.

What it does not do

  • It does not enforce policy.
  • It does not store prompts or assessment answers.
  • It does not replace gateways, IAM, policy engines, egress controls, or runtime controls.
  • It does not make compliance claims by itself.

Assessment use

Bring the framework into the work session.

An assistant can retrieve the right framework layer, ask first-pass governance questions, and point a team toward the right source documents.

1. Identify the delivery mode. AMS, waterfall, agile, product, platform, or mixed.
2. Route the problem. ContextOps, ContextBoundary, or Sthala; Griha as a worked example of governed capabilities in a running system.
3. Ask governance questions. Ownership, egress, runtime, approval, audit, and handover.
4. Keep enforcement deterministic. Gateways, IAM, policy engines, approval workflows, egress controls, and runtime controls enforce the rules.

Example questions

Use questions that route work, not vague prompts.

These examples are meant for an AI assistant connected to the MCP endpoint or given the framework documents directly.

Delivery

  • Which part of Context Stack applies to stale runbooks after AMS handover?
  • What should we assess first for AI context governance in a product team?
  • Which framework document should a delivery team read first?

Architecture

  • Is this a ContextOps issue, a ContextBoundary issue, or a Sthala issue?
  • Which governance questions should we ask before an AI workflow calls an external API?
  • Which controls must remain outside the model?

Endpoint

Read-only access.

The endpoint exposes resources, prompts, and guidance tools. It does not store organization-specific answers.

https://mcp.context-stack.org/mcp

JSON-RPC smoke test

curl -X POST https://mcp.context-stack.org/mcp \
  -H "content-type: application/json" \
  -d '{"jsonrpc":"2.0","id":1,"method":"tools/list"}'