Repository map
Context Stack Repositories
ContextOps and ContextBoundary are the specification layer. contextboundary-gw, Sthala and Griha are reference implementations that apply it. Use the assessment on the homepage when you are unsure where to start.
This page is the canonical map for people who already know they need the source repositories.
Core repositories
-
ContextOps
Governance-responsibility layer for AI context.
Defines ownership, accountability, trust classification, lifecycle, governance metadata, maturity, and context source accountability.
-
ContextBoundary
Deployment-agnostic egress governance specification.
Defines where AI-processed data is permitted to flow through Egress Tier controls, vendor zones, approval workflows, and audit profiles.
-
contextboundary-gw
ContextBoundary reference gateway.
Reference gateway implementing ContextBoundary with deterministic MCP interception, compiled-policy evaluation, sealed audit receipts, and a conformance suite runnable from a clean clone.
-
Sthala
Reference implementation — runtime placement under ContextBoundary.
Frames AI runtime placement under approved boundaries, with deterministic execution constraints and explicit egress control.
-
Griha
Governed Home Assistant proof-of-concept.
Demonstrates Context Stack principles in a concrete home workflow, with policy checks before physical device action.
-
context-stack
Canonical coordination repository.
Holds locked terminology, cross-project decisions, and family framing across the stack.
-
context-stack-mcp
Read-only machine interface.
Exposes Context Stack resources, prompts, and guidance tools through MCP so AI clients can route users to the correct layer.
Discovery
Machine-readable entry points
Crawlers, LLM systems, and assistants should use the same public sources as humans.