A set of templates — quality gates, skills, release scripts, CI — plus one skill that converts an existing repo into a standardized workspace agents can build, test, and ship behind enforced quality gates.
git clone https://github.com/Nimblesite/AgentPMOWorkflow.git && cd AgentPMOWorkflow && make setup
macOS, Linux & Windows · make setup auto-detects your OS · .NET & gh CLI auto-installed · full guide →
$ claude "run the agent-pmo skill" [agent-pmo] applying repo standards… ✓ Makefile 7 standard targets ✓ CI workflow PR gate + release ✓ quality gates lint · types · tests · coverage ✓ AGENTS.md agent instructions ✓ release script tag → build → deploy ✓ repo is now an Agent PMO workspace $ ▋
"Every repo is different — different structure, scripts, CI. Agents can't navigate chaos, so every project needs hand-holding and ships unverified code."
Inconsistent repos are the real bottleneck. When the build, test, and release commands differ everywhere, agents guess, and nothing guarantees the code is gated before it ships.
Agent PMO fixes that with templates and one skill. The skill rewrites an existing repo to a
single standard: one Makefile interface, enforced quality gates, agent instructions,
and a release script — so any agent already knows how to build, test, and ship it.
Run the skill once. The repo becomes easy to deploy, sits behind quality gates, and every agent works it the same way.
None of this is improvised. The standards here are evidence-based, and the research behind them — plus the broader strategy for deploying AI across an engineering team — is laid out in Nimblesite's guide, How to Deploy AI in Your Engineering Team. Agent PMO is one example of how a company can structure its repo templates for efficient, verifiable AI development.
Nimblesite helps engineering teams improve DevEx, standardise repos, add quality gates, and adopt AI coding workflows without losing control of review, testing, and delivery.
The PMO cycle runs continuously across every project simultaneously.
Dashboard scans all repos every 3 minutes. CI status, open PRs, uncommitted changes — everything at a glance.
You decide where to direct attention. Dashboard tells you — you choose. Minimal cognitive load, maximum leverage.
Drop agents into standardized repos. Same build targets, same CI, same lint commands. No setup. No babysitting.
TMC orchestrates agent coordination. Agents lock files, register intent, avoid conflicts. You watch the dashboard.
Code arrives production-ready. Lint, tests, coverage — all passed. You review intent, not whether it works.
Two components. One system. Infinitely scalable.
Scans every repo under ~/Documents/Code/ — CI status, open PRs, uncommitted changes, push status — and refreshes automatically every few minutes.
Portfolio-wide templates enforced by a Claude skill. Same build targets, same CI, same lint commands — every project, zero variation.
Lint. Type check. Format. Unit tests. Integration tests. Coverage thresholds — monotonically increasing. No soft-fail mode. Code ships clean.
No conductor — agents organize among themselves. Through Too Many Cooks, they register, lock files, broadcast plans, and message each other on one MCP server, so many work the same repo at once — one tree, no worktrees, no collisions.
Review every commit. Or every PR. Or just check the dashboard. You operate at the altitude that suits you — all options produce shipping code.
Public Nimblesite repos built on these standards. See all Nimblesite tools →
The coordination layer Agent PMO agents use. Its repo is managed by the same loop: agents register, lock files, publish plans, run standard targets, and ship MCP plus VS Code releases together.
Workspace command discovery. Agent PMO gives every agent the same command contract, coverage thresholds, and CI gate while TMC keeps extension and website work coordinated.
API testing CLI and editor tooling. Agent PMO keeps .NET, Rust, TypeScript, generated types, VSIX packaging, release binaries, and docs moving under one dashboard.
Schema DSL, code generator, web app, and VSIX. Agent PMO standardizes package tests, coverage ratchets, build checks, release checks, and multi-agent work across the repo.
.NET data toolkit. Agent PMO coordinates LQL, migrations, sync, generator packages, specs, coverage thresholds, and CI so separate agents can own different product surfaces.
Live duplicate-code analysis for humans and agents. Agent PMO uses it as a quality gate, and Deslop itself is built with PMO standards, TMC coordination, and Shipwright release verification.
Version-matched binaries and IDE extensions. Agent PMO drives the specs, release tooling, coverage gates, and downstream adoption checklists that keep packaged tools honest.
Python language server and toolchain. Agent PMO coordinates the Rust core, editor extensions, website, Deslop checks, Shipwright manifests, coverage, and multi-agent PR flow.
Use Nimblesite to audit your current workflow, remove DevEx bottlenecks, and install the repo standards, dashboards, and agent coordination needed for safe AI-assisted development.
Lint, type check, format, tests, coverage — all enforced by CI. No soft-fail mode. By the time code reaches you, every automated check has passed. You're the final gate: reviewing intent and architecture, not whether it compiles.
All options produce shipping code. You decide how close to stay.
Diffs are clean. Tests passed. Every commit is a complete, verified unit of work.
Every PR is already production-ready. Your review is about the what, not the whether it works.
You're still shipping — just operating at a different altitude. The dashboard tells you what's happening.
Every spec section gets a unique ID — [AUTH-LOGIN], [AUTH-TOKEN-VERIFY].
Code references the spec it implements. Tests reference the spec they verify.
PRs list the spec IDs they address. Full bidirectional traceability across the entire portfolio.
