magus v0.4.3 is out. See what's new
¶ View markdown source · ✎ Suggest an edit
2 min read

architecture-unbriefed

A deny rule by default: it refuses an agent's next call after a structure question, or a new directory, before the architecture skill, and names what to run instead.

What it catches

An agent's next call after a structure question, or a new directory, before the architecture skill.

Why

A structure question is answered from the workspace's own edges: what imports what, where a cycle closes, which layer a package sits in, how far a change reaches. An agent reading files one at a time answers it from the files it happened to open, and proposes a boundary nobody needed. The magus-architecture-review skill is how the graph gets asked. Two acts arm it. A message the person submits with structure vocabulary (architecture, boundaries, layering, coupling, imports, a new package, where something belongs, blast radius) marks the session, and the agent's next graded call (a shell line, a read, a search, an edit, a magus tool call) waits on the load. A write that creates a new directory waits on it too, since that write draws a boundary whatever was asked. A load counts for the agent that made it: a subagent in that session loads the skill itself, and a parent's load does not brief its child. Load Skill(magus-architecture-review), or its -full twin, once and the agent's later calls pass. It stands down on a host whose wiring does not report skill loads, where nothing could ever clear it; the new-directory advisory still speaks there. It also stands down for an agent spawned as a shipped read-only type (magus-scout), which has no Skill tool: a deny it cannot clear is a trap, and such an agent changes nothing.

Default and override

By default this rule takes the decision deny. A workspace sets it by name, in its root magusfile, to deny, advise or off:

magus\guard.builtins({"architecture-unbriefed": "advise"})

A loosening takes effect once it is committed; a tightening applies at once.

Seeing it

A verdict names its rule in brackets, which is how you got here:

deny [architecture-unbriefed]: ...

A workspace that sets this rule to advise sees advise [architecture-unbriefed]: ... instead.

magus describe rule architecture-unbriefed prints the same entry at a terminal, and magus describe rules lists every rule this workspace enforces.

See also

guardrulesarchitecture-unbriefeddeny
Last updated (c6a327d5)
Earlier changes on this page (2)

Full history ↗ · Blame source ↗

Glossary

Workspace

The magus root directory that owns a set of projects and shared config; the unit magus operates over. Workspace in full.

Magusfile

The magusfile.buzz that declares a project's targets (as export funs) and binds its spells. Magusfile in full.

Op

A single tool-native command a target composes (long form: operation); the middle of the work hierarchy (Spell to Op to Target). Op in full.

Buzz

The language magusfiles are written in (the .buzz engine). Buzz in full.

CI

An ordinary target you compose yourself with ctx.needs; magus does not hardcode its stages. It is the anchor magus affected ci keys off, and ci strips the rw charm. CI in full.

Agent

A model working through an agent host such as Claude Code, Codex, Cursor or OpenCode. magus treats an agent as one more caller of its CLI, MCP tools and console, held to the same checks as a person. Agent in full.

Tool

An action a model can call through its host: a name, a description, a typed input and a result. magus serves its tools over MCP, each a read of the workspace model or a run of the CLI it already has. Tool in full.

Skill

A packaged set of instructions an agent host loads when a task matches its description: a directory holding a SKILL.md. magus ships its magus-* skills in the binary and writes them with magus agent install. Skill in full.

Session

An agent host's conversation, by the id the host delivers to its hooks. magus never mints one: a record with no session is unattributed, and the OS user it carries says whose account ran it. Session in full.

Run

One target executing under one magus invocation, such as magus run test web or magus affected ci. A run keeps its captured output behind an output reference. Every magus run is a run whether or not any job asked for it; see Job for how the two relate. Run in full.

Advisor

One read-only check from the advice suite: it reads the changeset through magus and writes one titled section of findings. The same advisors run as a pull request comment in CI and inside magus diff --impact locally. Advisor in full.

Conventions

This page uses none of the site's convention markers. The full set is on the conventions page.