An agent organization structured as a company — 15+ departments, 125+ skills, each independently installable, citing the standards and regulators that settle the question. Runs in Claude Code and ChatGPT.
Agent surfaces
The write-surface map for this repository. Every tracked path has exactly one owner and no two
owners claim the same path. agent-guard check verifies both, and CI runs it on every push.
Read executive:agent-hierarchy for the method. The short version: split by exclusive write
surface, not by topic, because a topic split has no checkable boundary — two agents working on "SEO"
and "UI" both end up in the same token file, and neither is wrong.
Classes
- builder — edits inside exactly one exclusive surface, never commits.
- reviewer — permanently read-only, holds no write surface, can always run in parallel.
The orchestrator is neither: it owns no surface and is the sole committer.
Authority
The class and the surface together answer where an agent may write. They have never answered
whether that write may land without a decision, so in practice that was settled per dispatch, from
memory, by whoever was driving. Authority states it once, in the same row, and check verifies it.
- autonomous — dispatch it and take the result. The surface is the only gate needed.
- proposes — it may do the work; the orchestrator surfaces the diff before landing it.
- escalates — do not dispatch it unasked. The work itself is the decision.
Two things the check enforces, both because the row would otherwise read as governed while governing nothing: a reviewer may not be gated, because it holds no write surface to gate; and a gated builder must actually own a surface.
Almost every row here is autonomous, and that is the honest answer rather than a placeholder — a
department writes only inside its own plugin directory, where the worst case is a bad skill in one
department. The column exists because the one exception is real, and because a repository adopting
this map will have more of them than this one does.
Roster
# id class status authority
executive builder installed autonomous
technology builder installed autonomous
product builder installed autonomous
marketing builder installed autonomous
demand-generation builder installed autonomous
revenue builder installed autonomous
finance builder installed autonomous
operations builder installed autonomous
people builder installed autonomous
legal-risk builder installed autonomous
customer-experience builder installed autonomous
data-analytics builder installed autonomous
corporate-strategy builder installed autonomous
security builder installed autonomous
it-operations builder installed autonomous
pmo builder installed autonomous
verticals builder installed proposes
sources builder installed autonomous
repo-meta builder installed proposes
legal-risk-review reviewer installed autonomous
security-review reviewer installed autonomous
repo-meta is the exception because of what it owns: the CI workflows, the check scripts, the
generators every document is built from, and this map. A change inside plugins/finance/** is
wrong in one department. A change to scripts/check-all.sh can make every other check stop
reporting, and nothing downstream would fail to say so.
verticals is the second, for the same reason in a different direction. It owns the per-vertical
configs and content that scripts/build-vertical.py emits standalone repositories from, so a change
inside verticals/** does not land in this repository — it lands in a distributable artifact
carrying somebody's industry advice, and the emit is one-way, so nothing downstream can correct it
locally. It owns the input rather than the generator; the generator is a script and belongs to
repo-meta with the others.
Charters live in .claude/agents/, one per installed row. A row marked installed without a
charter, or a charter without a row, fails the check — the two cannot drift apart silently.
The column may be omitted; an omitted authority means autonomous, and check reports which rows
defaulted so that a map which never considered the question is distinguishable from one that
answered it.
Surfaces
plugins/executive/**
plugins/technology/**
plugins/product/**
plugins/marketing/**
plugins/demand-generation/**
plugins/revenue/**
plugins/finance/**
plugins/operations/**
plugins/people/**
plugins/legal-risk/**
plugins/customer-experience/**
plugins/data-analytics/**
plugins/corporate-strategy/**
plugins/it-operations/**
plugins/pmo/**
plugins/security/**
verticals/**
sources/**
LICENSE
.gitignore
AGENTS.md
.agents/**
CONTRIBUTING.md
.gitattributes
docs/**
scripts/**
.github/**
.claude/**
.claude-plugin/**
README.md
Reviewers declare no surface. That is structural rather than a promise: check fails if a reviewer
claims one.
Reviewer independence
legal-risk appears twice on purpose. As a department it owns plugins/legal-risk/** like any
other builder. As legal-risk-review it is reviewer-class: it reviews what other departments
commit to, holds no surface in that capacity, and its findings are not overrulable by the department
under review. Disagreement escalates to the Chief Executive rather than resolving inside the
reviewed department. See D13.
security mirrors legal-risk: a builder owning plugins/security/**, and separately
security-review, reviewer-class over what other departments build. Its blocking findings are not
overrulable by the department under review, which is why the CISO reports independently rather than
under the CTO (D9, D13).
Rules
- One owner per path. A new department adds its roster row, its surface block, and its charter in the same change, or the check fails.
- Never remove a surface because it looks unused. Deprecate, announce, then remove.
- Cross-surface work moves as one coordinated change, never as two independent ones.