miklschmidt/archboard / skills/archboard
archboard
VerifiedRead, author, and visualize local architecture with the archboard CLI. Always use in repositories with an .archboard/vault directory: when planning, implementing, explaining, visualizing, refactoring, or improving the codebase or code in general. Load this skill when loading, reading, or using improve-codebase-architecture. Also use for architecture questions, request flows, proposals, and explicit migration of existing architecture vaults.
Archboard
Use the CLI to read architecture, author changes, and produce diagrams.
Model every flowchart and sequence chart as a flow, and draw it with model render --flow <id> --grammar flowchart or --grammar sequence.
Never write a flowchart or sequence chart in Mermaid.
A vault is the directory that holds one model.
A reading selects what you see without copying architecture.
Choose the work
Before the first architecture operation, read canonical models for discovery, reads, writes, and history.
For LEGACY_AUTHORITY or requested migration, follow legacy migration.
Before describing implemented architecture or authoring changes, read authoring for evidence and modeling rules.
Before preparing a request, read schemas.
For design decisions or refactors, read the design lens.
For improve-codebase-architecture, follow the integration workflow.
For narrative authoring, playback, or recording, read canonical walkthroughs.
For installation or updates, read the installation guide.
For missing language servers or more precise source evidence, read setup.
Make the vault
If the Git checkout has no .archboard/vault, a command refuses with VAULT_NOT_FOUND.
If the refusal names another vault that holds the models of this repository, pass that vault with -C.
Otherwise, make the vault with model init in the checkout before any other operation:
archboard model init --name <name> --operation-id initialize-<name> --writer agent --doing 'Initialize architecture'
Use the name of the repository directory as the model name, unless the user names the model.
The command makes <root>/.archboard/vault with the default vocabulary and an empty planning model.
It registers the checkout, so that a browser lists it. It needs no server.
Tell the user that you made the vault, and where.
Then configure the repository's language servers as setup says.
Tell the user to commit .archboard/language-servers.json with the vault.
Complete the intent
Read the resulting architecture.
Compare its stable subjects with the intended change.
Make sure that the requested scope, flow, and every narrative beat remain available.
Run affected semantic, code, and embed inspection through archboard check.
Report concrete changes, refusals, unavailable source, and recovery work.
Claim frontend acceptance, adoption, publication, or migration only after its operation succeeds.
Source evidence catalogue
Use each feature when the source or requested design establishes it.
| Row | Canonical evidence |
|---|---|
external | Mark a module outside the described ownership boundary. |
binding | Bind a module to the source that implements its responsibility. |
containment | State parent identities without a fixed level. |
relationship | Author each relationship on its source module. |
traffic | State meaningful effective traffic properties. |
emphasis | Mark the main exchange when it helps the reading. |
repeat | Record repeated exchanges on authored steps. |
note | Attach relevant evidence or constraints to a step. |
groups | Apply configured group identities. |
flow | Author a complete, described exchange in a collection. |
view | Select a content-free reading or imported bookmark. |
walkthrough | Author ordered beats with exact reading targets. |
drillDown | Follow an explicit model reference or recursive module scope. |
description | Explain responsibilities and significant architectural decisions. |
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 470k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |