Skip to content
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
4 changes: 4 additions & 0 deletions CONTRIBUTING.md
Original file line number Diff line number Diff line change
Expand Up @@ -78,6 +78,10 @@ index and drill down to the right page. Three conventions make that work.
`src/zensical.toml`, and fills every index in place. CI runs the same script with `-Check`
and fails if an index is out of date.

Choose a page's home by subject first, then by the artifact that answers the reader's
question. The [Documentation Model](https://msxorg.github.io/docs/Ways-of-Working/Documentation-Model/#architecture-and-classification)
defines those artifact tiers and the deliberate use of minimal OKF-style front matter.

Links are validated the same way: `.github/scripts/Test-DocumentationLink.ps1` checks that
every relative link and heading anchor across the docs resolves, and
`.github/scripts/Test-CrossRepositoryLink.ps1` resolves every link into another MSX
Expand Down
Original file line number Diff line number Diff line change
@@ -0,0 +1,48 @@
---
title: Document architecture and memory boundaries
description: The topic-and-artifact documentation model, minimal OKF-style metadata, and separate durable memory repository.
---

# Document architecture and memory boundaries

## Context

MSX documentation is navigable cheaply by humans and agents, agent
configuration points to canonical knowledge, and durable lessons have a shared
home. The architecture establishes those boundaries without duplicating
documentation across classifications or repositories.

## Decision

MSX uses the [Documentation Model](../../../Ways-of-Working/Documentation-Model.md)
as the documentation architecture:

- Paths are topic- and scope-oriented, and every area is navigated through
`index.md`.
- The spec, design, guide, reference, decision-record, and research artifact
tiers classify content by the reader's need.
- Pages use the minimal OKF-style model: Markdown, YAML `title` and
`description` front matter, one primary concept per page, and stable paths.
- `MSXOrg/docs` remains the reviewed, pull-request-only canonical knowledge
base. Durable working knowledge belongs in the separate private
`MSXOrg/memory` repository; session notes remain local and ignored.

MSX does not add Diataxis quadrant directories or `diataxis` metadata. The
artifact tiers already route the same reader needs while keeping related
subject matter together. MSX also does not adopt strict OKF conformance fields
or per-area `log.md` files: the only metadata queried by navigation is title
and description, and Git history is the authoritative changelog.

## Consequences

Contributors file a page by subject and artifact tier, then run the index
generator. Readers and agents traverse the same indexes without loading an
unrelated quadrant. Tooling validates the metadata it consumes and validates
links, while review keeps page boundaries and cross-references coherent.

Agents propose changes to canonical documentation through pull requests.
Agents commit durable, factual lessons to `MSXOrg/memory` under its
[memory-writing rules](../design.md#memory-writing-rules); they never use this
public documentation repository as a low-ceremony dump.

Changes to these boundaries use a new decision record that supersedes this one.
17 changes: 17 additions & 0 deletions src/docs/Capabilities/agentic-development/decisions/index.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,17 @@
---
title: Agentic Development decisions
description: Immutable records of one-way-door choices in the agentic development framework.
---

# Agentic Development decisions

These records explain choices that constrain the framework. A later decision
supersedes an earlier one; it does not rewrite it.

<!-- INDEX:START -->

| Page | Description |
| --- | --- |
| [Document architecture and memory boundaries](document-architecture-and-memory-boundaries.md) | The topic-and-artifact documentation model, minimal OKF-style metadata, and separate durable memory repository. |

<!-- INDEX:END -->
1 change: 1 addition & 0 deletions src/docs/Capabilities/agentic-development/index.md
Original file line number Diff line number Diff line change
Expand Up @@ -22,5 +22,6 @@ A repository adopts the framework by carrying a short router and the client rout
| [Agent Interaction](agent-interaction.md) | How humans and agents coordinate through issues, labels, and pull requests, and why intent and implementation are kept in separate artifacts. |
| [Advisory Agents](advisory-agents.md) | The pattern for automation that analyses work and publishes its conclusion as advice, without deciding, relabelling, or committing. |
| [Conformance](conformance.md) | What a repository must provide to be conformant with the agentic development framework, what it may add, and the duplication checks that keep the router thin. |
| [Agentic Development decisions](decisions/index.md) | Immutable records of one-way-door choices in the agentic development framework. |

<!-- INDEX:END -->
18 changes: 18 additions & 0 deletions src/docs/Capabilities/deployment/designs/index.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,18 @@
---
title: Deployment designs
description: How the deployment contract is delivered for each service-provider and CI/CD-platform combination.
---

# Deployment designs

Each page in this section delivers the [Deployment spec](../spec.md) for one
service-provider and CI/CD-platform combination. Add a design here without
changing the technology-agnostic contract.

<!-- INDEX:START -->

| Page | Description |
| --- | --- |
| [Deploying Azure from GitHub](azure-from-github.md) | How the deployment spec is delivered with Azure as the service provider and GitHub as the CI/CD platform, using Terraform to compute and apply the approved effect. |

<!-- INDEX:END -->
2 changes: 1 addition & 1 deletion src/docs/Capabilities/deployment/index.md
Original file line number Diff line number Diff line change
Expand Up @@ -29,4 +29,4 @@ combination adds a design; it never changes the spec.

| Design | Service provider | CI/CD platform | Description |
| --- | --- | --- | --- |
| [Deploying Azure from GitHub](designs/azure-from-github.md) | Azure | GitHub | GitHub Actions and Terraform deploy Azure and Entra resources with passwordless identity, approving the code change together with its per-environment effect. |
| [Designs](designs/index.md) | — | — | Deployment designs, one for each service-provider and CI/CD-platform combination. |
26 changes: 26 additions & 0 deletions src/docs/Ways-of-Working/Documentation-Model.md
Original file line number Diff line number Diff line change
Expand Up @@ -176,6 +176,32 @@ being long.
describe. Laziness is a design constraint — the less a reader must travel, the
more they actually read.

## Architecture and classification

The documentation tree is organized **by scope and subject**, then by the
artifact that answers the reader's question. A reader starts at the root
`index.md`, chooses the relevant topic, and follows that topic's index inward.
This keeps related material together without requiring a reader to decide
whether they need a tutorial, guide, reference, or explanation before they can
find the subject.

The artifact tiers provide the second classification:

| Reader need | Owning artifact |
| --- | --- |
| Understand why a capability exists and what it guarantees | Spec |
| Understand how the capability is delivered | Design |
| Perform a task against a capability | Guide |
| Look up stable facts or values | Reference |
| Understand a one-way-door choice | Decision record |
| Revisit point-in-time findings | Research |

This maps the useful reader needs behind Diataxis without imposing a second,
parallel taxonomy in paths or front matter. A page has one home under its
subject and links to adjacent artifacts rather than being duplicated across
quadrants. [Document architecture and memory boundaries](../Capabilities/agentic-development/decisions/document-architecture-and-memory-boundaries.md)
records that decision and the corresponding OKF and memory boundaries.

## For humans and agents

The same pages serve both. A contributor reads the index, follows the
Expand Down
9 changes: 8 additions & 1 deletion src/zensical.toml
Original file line number Diff line number Diff line change
Expand Up @@ -152,7 +152,10 @@ nav = [
{"Deployment" = [
"Capabilities/deployment/index.md",
{"Spec" = "Capabilities/deployment/spec.md"},
{"Deploying Azure from GitHub" = "Capabilities/deployment/designs/azure-from-github.md"},
{"Designs" = [
"Capabilities/deployment/designs/index.md",
{"Deploying Azure from GitHub" = "Capabilities/deployment/designs/azure-from-github.md"},
]},
]},
{"VS Code Extension Framework" = [
"Capabilities/vscode-extension-framework/index.md",
Expand All @@ -175,6 +178,10 @@ nav = [
{"Agent Interaction" = "Capabilities/agentic-development/agent-interaction.md"},
{"Advisory Agents" = "Capabilities/agentic-development/advisory-agents.md"},
{"Conformance" = "Capabilities/agentic-development/conformance.md"},
{"Decisions" = [
"Capabilities/agentic-development/decisions/index.md",
{"Document Architecture and Memory Boundaries" = "Capabilities/agentic-development/decisions/document-architecture-and-memory-boundaries.md"},
]},
]},
]},
{"Dictionary" = "Dictionary/index.md"},
Expand Down
Loading