Use this map to choose the smallest useful maintainer document. Detailed
package behavior belongs in each skills\martix-* or plugins\martix-* package.
.github\copilot-instructions.mdis the concise repository policy for agents..github\instructions\*.instructions.mdapplies automatically by file scope.- Package
SKILL.md,AGENTS.md, andREADME.mdare the active source for package behavior.
The unified documentation tree is knowledge. Each topic may contain:
README.mdfor scope and navigation.resources.mdfor canonical upstream links.knowledge.mdfor the current synthesis or decision layer.research/for dated, source-backed snapshots.
Start with the knowledge map, then use the topic map for the relevant domain:
| Topic | Scope |
|---|---|
| AI-assisted development | AI usage, hooks, LLMs, plugins, skills, tools, and Copilot. |
| Matt Pocock | Matt Skills, Sandcastle, and related workflow research. |
| Repository | Repository architecture, maintenance, and Git workflow research. |
Research snapshots are non-normative. Revalidate volatile upstream claims before moving them into active guidance, and add a new dated snapshot rather than silently rewriting an older baseline.
| Guide | Use when |
|---|---|
| Repository knowledge | Understanding layout, package rules, installs, and current architecture. |
| Execution and routing | Choosing model tier, routing surface, token budget, task classification, or worktree safety. |
| Parallel worktree guidance | Splitting package work across agents or branches. |
| Skills installation | Importing external skills into a target repository. |
| MX configuration | Configuring the central mx.config.json, schema, presets, templates, and workflow behavior. |
| Custom AI artifact rules | Creating or updating skills, plugins, agents, prompts, hooks, metadata, templates, or evals. |
| Custom AI artifact resources | Primary-source links for artifact authoring and host behavior. |
| Plugin bundle strategy | Deciding standalone skill versus plugin bundle boundaries. |
| Repository roadmap | Tracking planned repository, package, validation, and registry work. |
| Skill portfolio coordination | Coordinating related skills and handoffs. |
| Recommended skills research | Reviewing dated external skill or plugin candidates. |
| Skill evaluation research | Reviewing evidence and quality criteria for skill evaluations. |
| Plugin layout policy | Checking machine-readable package layout assumptions. |
The old root-level guide paths remain short compatibility facades for existing links. New links should use the canonical locations above.
Read these on demand when changing the related package or reviewing its history. Package behavior remains canonical in the package itself.
| Folder | Scope |
|---|---|
| agents | Issue tracking, triage, and agent-domain operating guidance. |
| martix-dotnet-csharp | .NET/C# comparisons and plans. |
| martix-fastendpoints | FastEndpoints improvement planning. |
| martix-fluentvalidation | FluentValidation improvement planning. |
| martix-fluent-ui | Fluent UI research, evidence, and skill blueprint. |
| martix-essl | Czech eSSL source research and compliance maps. |
| martix-markdown | Markdown automation and package split notes. |
| martix-csharp | Earlier standalone C# skill planning artifacts. |
| martix-power-platform | Power Platform and custom connector source research. |
| martix-typescript | TypeScript skill planning and source research. |
- Use package
rules\for enforceable guidance loaded by topic. - Use package
references\for maps, recipes, source indexes, and anti-patterns. - Use package
templates\,assets\, andevals\when changing generation or validation behavior.
| Task | Start with |
|---|---|
| Understand the repository | Repository knowledge |
| Find current knowledge or research | Knowledge map |
| Create or update a skill or plugin artifact | Custom AI artifact rules |
| Choose skill versus plugin ownership | Plugin bundle strategy |
| Plan parallel package work | Execution and routing |
| Update one package | Package README.md, SKILL.md, and AGENTS.md |