Gives a team and its AI agents shared ways of working for a project, covering planning, ticket tracking, model choice, testing, independent review and release checks. Each project gets one operating contract that says who decides and what must pass before delivery.
Claude Code
> /plugin marketplace add jpantsjoha/ai-native-developer-experience> /plugin install join-the-team@join-the-team-marketplace
Codex
$ npx skills add jpantsjoha/ai-native-developer-experience -a codex
Cursor
$ npx skills add jpantsjoha/ai-native-developer-experience -a cursor
Gemini CLI
$ npx skills add jpantsjoha/ai-native-developer-experience -a gemini-cli
GitHub Copilot
$ copilot plugin marketplace add jpantsjoha/ai-native-developer-experience$ copilot plugin install join-the-team@join-the-team-marketplace
OpenCode
$ npx skills add jpantsjoha/ai-native-developer-experience -a opencode
Windsurf
$ npx skills add jpantsjoha/ai-native-developer-experience -a windsurf
Listed on Agent Plugins Directory
Shared ways of working for you, your team and your AI agents — from a new idea to a change running live.
join-the-team brings planning, delivery and ticket management, model selection,
testing, independent review and release practices into one reusable plugin. Install it
for a project or across your own projects, then give each project a shared operating
contract: what you are building, who owns the decisions, how work is tracked and what
must pass before delivery.
The aim is reliable delivery with less repeated setup, lost context and avoidable rework. The plugin supplies skills, templates and executable checks; your project supplies its goals, people, tools and authority. Installation makes those capabilities available. Adoption makes them part of how the team works.
| From intent to live | What the harness provides |
|---|---|
| Agree the work | Vision, requirements and acceptance criteria before implementation |
| Plan and track it | Owned tasks, dependencies, tickets, roadmap and current status |
| Choose how to execute | Skills, model-selection guidance, bounded delegation and cost checks |
| Build and verify | Isolated work, executable tests, validation and independent review |
| Release and operate | Readiness checks, explicit promotion authority, rollback and observation |
| Improve the practice | Reconcile decisions and outcomes into existing project records |
Use it alone or with a team. Team members share the project’s contract and records; their assistants can use different supported hosts. Platform adapters handle discovery and invocation; detailed host support is documented below. The plugin does not replace your issue tracker, provision CI, deploy an application or guarantee a performance gain.
Start here: install for your host, then bootstrap a project. Existing projects keep their records; bootstrap reports conflicting contracts before writing.
Current source status · Roadmap · Releases · Agent Plugins Directory
By hand — clone the complete repository to work on this harness. To adopt it in
another project, copy skills/ to .agents/skills/ and follow
BOOTSTRAP.md to generate that project’s adapters and operating contract.
The root adapters and active routing pilot in this repository are maintainer settings;
do not copy them into an adopter project.
git clone https://github.com/jpantsjoha/ai-native-developer-experience
As a plugin — the harness ships as join-the-team: one canonical skill
set, thin per-harness adapters, drift-checked in CI. It is listed on
Agent Plugins Directory;
install straight from this repository into whichever coding assistant you run:
Claude Code — choose user scope for your projects or project scope for a shared repository. See the scope and team setup guide.
claude plugin marketplace add jpantsjoha/ai-native-developer-experience
claude plugin install join-the-team@join-the-team-marketplace --scope user
Kimi Code
/plugins install https://github.com/jpantsjoha/ai-native-developer-experience
Antigravity (Gemini)
agy plugin install https://github.com/jpantsjoha/ai-native-developer-experience
Codex — use the Plugins interface when available, or install the skills in
.agents/skills/ for a project or ~/.agents/skills/ for your user. Keep each skill’s
supporting files together and avoid competing copies. Follow the
Codex guide for discovery and verification.
Global/user scope applies to one user’s host setup; it does not install the plugin for every teammate. Each collaborator needs access to the selected skill version. Commit the adopted project contract and agreed host configuration so people share the same ways of working, while credentials and personal settings stay private.
Hooks are optional host conveniences. Without them, invoke using-the-harness and
operating-model-bootstrap from the consumer project; verify its generated agent
adapter and profile are loaded. Installation alone does not adopt project controls.
Full per-harness detail (session-start hooks, verification, update path): Claude Code · Kimi Code · Codex · Antigravity
Antigravity is a first-class surface: this harness was built and battle-tested on
Google Cloud's agent stack, and ships cloud guardrails (cloud-expert, with GCP, AWS, Azure and Alibaba Cloud references) plus adk-expert and
mcp-server-scaffold alongside the vendor-neutral contract.
The plugin composes with — never duplicates — companion skill plugins; see INTEGRATIONS.md.
The golden thread connects evolving intent to versioned requirements and business
journeys, ADRs, roadmap outcomes, GitHub milestones/issues, acceptance checks and releases.
product-owner maintains intent and priorities; cxo-oversight challenges material
value/risk decisions; test-strategy maps acceptance to the real user path.
delivery-orchestrator checks in-flight work first, maintains tickets and reconciles
records through reviewed delivery and safe branch/worktree cleanup. See the
portable workflow.
Root AGENTS.md, CLAUDE.md and GEMINI.md share identical lean instructions. Session hooks
inject a compact pointer by default, even for adopted projects; detailed roles and
procedures load when needed. Set JOIN_THE_TEAM_ORIENTATION=full for full orientation.
Model choice follows guarded routing, including qualification,
risk floors and retry/integration costs; a task's T-shirt size does not select its model.
These are maintained agent workflows, not an automatic requirements synchroniser or
measured efficiency claim. Existing adopters must deliberately reconcile their local
adapters/workflow when upgrading; preserved project files are not silently overwritten.
The library contains 25 skills. After installing, verify discovery and project adoption:
which skills are available?
You should see using-the-harness plus the cloud and gate skills (cloud-expert,
adversarial-gate, pr-reviewer). Then run the built-in validator against a
bootstrapped repository using /join-the-team:validate where available, or the
operating-model-bootstrap skill on skill-only hosts:
/join-the-team:validate
Expect PASS operating-model validation. A seed profile may pass with warnings — that
is expected before the profile is promoted to active.
The harness is not a set of commands you memorise. Skills trigger on intent; three slash commands cover the lifecycle.
On hosts that expose the plugin commands:
/join-the-team:bootstrap # seed the operating model (dry-run first, never overwrites)
/join-the-team:init # record the humans: roles, authority, escalation channels
/join-the-team:validate # check adapter drift, adoption state, evidence binding
On skill-only hosts, ask operating-model-bootstrap to run the same sequence. It finds
scripts in the selected installation and writes only the consumer project’s seed.
Typical sessions — say the intent, and the router picks the skill:
| You say | What fires | What you get |
|---|---|---|
| "Design the ingestion service" | spec-first-delivery → the-architect | Spec and ADR before code |
| "Is this safe to ship?" | release-readiness | Go/no-go against failure modes and rollback |
| "Review this PR" | pr-reviewer | Structured verdict with findings |
| "How would this break?" | adversarial-gate | Red-team pass argued against your own approach |
| "We're on GCP, multi-tenant" | cloud-expert + governance-guardrail | IAM, residency and policy guardrails |
| "Where do things stand?" | sitrep | Status, blockers, next actions |
| "Get it done, I'll be back later" | get-it-done | Pursues the agreed goal to a merged outcome; banks decisions instead of blocking |
| "The Mac is slow / the tests left Chrome open" | system-performance-guardrail | Snapshot, owner attribution, bounded cleanup, before/after numbers |
| "Is everything live? Any errors this week?" | cloud-checkup | Read-only GCP checkup: one status table, one OVERALL: line, in session or on a schedule |
| "Is the page fast enough? Why is this route slow?" | frontend-performance-audit | Core Web Vitals and traces from the real authenticated page, not an estimate |
Worked example — a risk-touching change:
You: We need to let tenants export their own audit logs.
Agent: [spec-first-delivery] Spec before code. Who is the actor, what is out of scope?
[governance-guardrail] Multi-tenant export is R2 — data-boundary check required.
[the-architect] ADR-00N drafted: signed URLs vs streamed export, trade-offs recorded.
[adversarial-gate] How would I break this? Tenant A requesting tenant B's logs.
→ Named human approves the R2 classification before implementation starts.
The point is the routing, not the vocabulary: you describe intent, the contract decides which gate applies.
If you work across several repositories, DevDash is an optional native macOS desktop app for seeing which projects need attention. It collects delivery signals from GitHub and git, shows available signals and when they were checked, and keeps missing data visibly missing. Displayed metrics are deterministic. Optional AI proposes short explanations from available status and roadmap documents; it never supplies the displayed numbers. The core needs no AI subscription.
Used alongside join-the-team, clear status and roadmap records give DevDash more context
for the work behind those signals. The harness is not required: ordinary GitHub repositories
still provide available API and git signals, with less delivery context where documents are
missing. See the four delivery records for the shared practice.
DevDash repository (mac-control) — private preview, access required.
DevDash is currently an internal MVP, with no public signed download or release. The link is
accessible only to people granted repository access. Local AI is experimental and has not
completed qualification.
The skills are discoverable after installation. Adoption records the project’s controls before consequential work. Three seams are worth knowing:
| Seam | File | What it controls |
|---|---|---|
| Operating contract | CLAUDE.md / GEMINI.md / AGENTS.md | Always-on rules the agent loads each session |
| Project profile | PROJECT-OPERATING-PROFILE.md (generated by bootstrap) | Risk tiers, named authority, escalation, definition of done |
| Data seams | .agents/mcp_config.json | Governed MCP access — a template, not live config |
Adjust the risk tiers and authority model to your team; keep one coherent contract while you do. The MCP file ships as an example: copy an entry, rename it, point it at a source you control. See
Stars on GitHub
A development methodology for coding agents that makes the agent brainstorm before building, follow plans, run parallel work and finish branches in a disciplined way. You get fewer skipped steps and more consistent results from your coding assistant.
@obraHelps you take a change from a rough idea or bug report through to a tested, reviewed result, with each step handled by a focused skill. Its skills cover pinning down domain terms, designing module interfaces, diagnosing hard bugs, and reviewing changes against your repo's coding standards.
@mattpocockGives your coding agent a planning, testing and review routine to follow on every task. It plans before building, checks changes with tests, has a fresh reviewer look for regressions, and saves repeated wins as reusable skills.
@affaan-mDeveloper tools for working on the React codebase itself. Provides skills for extracting error messages, managing and debugging feature flags across release channels, running Flow type checks, and fixing lint and formatting issues before committing.
@facebookGives a team and its AI agents shared ways of working for a project, covering planning, ticket tracking, model choice, testing, independent review and release checks. Each project gets one operating contract that says who decides and what must pass before delivery. Install it with the command on its page. It includes adk-expert, adversarial-gate, cloud-checkup. Its skills do not fire on their own yet. Free and open source.
Is this plugin yours?
Claim it with GitHubSubmit a pluginPromote it