version: "1.0.1" name: onboard description: Use to bring a project into the DAE methodology, or to check an onboarded project for gaps. Triggers — "/engineer.onboard", "onboard this project", "set up DAE here", "adopt the methodology", or when a DAE skill fails because no manifest exists.

onboard

Checkpoint 0 — the DAE adoption ceremony. Establishes the charter, manifest, storage layout, and tracker. Project-scope, run once. Every other DAE skill depends on what it produces.

The goal. Onboarding a project to DAE succeeds when there is a clear path to full ATDD coverage of every feature — existing and new. A new feature is born covered by going through the pipeline. An existing feature is covered retroactively.

Onboarding is discovery and goal-setting — not the ATDD adoption itself. It discovers what's there (documented and undocumented), triages it by importance, assigns each feature a status, and produces a consolidation backlog. Bringing any one feature to full ATDD coverage is a follow-up task per feature — bounded, automatable, and a good candidate for remote-agent dispatch. Onboarding sets the path; it does not walk it.

A feature is fully ATDD-covered when its folder has feature.md, acs.md, spec.md (+ .build/spec.json IR), and generated acceptance tests that pass against the code.

When to use

• No `.engineer/manifest.yml` → full onboard (Steps 1–11)
• Manifest exists → gap-check mode (validate, report gaps, don't re-onboard)

Not for: starting a feature (discuss / feature-init, after onboard); changing an existing charter (edit it directly, PR'd).

Human-decision checkpoints

Onboarding is a ceremony, not a mechanical scaffold. Two of its outputs are design decisions reserved for the human — the agent drafts, the human decides:

• The charter (Step 3) — architecture, conventions, scope, quality and autonomy stance.
• The tracking decision (Step 5) — which tracker the project uses.

Pre-filling from an existing codebase is encouraged. Rubber-stamping is not. Onboarding does NOT complete until the human has explicitly signed off on the charter and chosen the tracker — exactly as plan does for architecture (agent proposes, human confirms before proceeding). If the human is not available to decide, stop and emit a handoff with human_action_needed: decision — do not auto-decide and move on.

Workflow (full onboard)

Before the steps below, create one TodoWrite todo per workflow step (the full list up front, as a roadmap) — see ${CLAUDE_PLUGIN_ROOT}/references/progress-indicator.md, Indicator 2. onboard is project-scope and has no feature folder, so it does not show the pipeline breadcrumb.

1. Repo topology — ask single- vs multi-repo. Set methodology_root (and repos[] for multi-repo).
2. Discover validation infrastructure. Before drafting the charter, probe what's available — the findings inform the autonomy stance.

• LSP probe. Walk the repo for language signals (file extensions; package files: requirements.txt, pyproject.toml, package.json, Cargo.toml, go.mod, pom.xml, *.csproj, …). Aggregate primary languages. Inspect the agent's available tool list for an LSP capability (per ${CLAUDE_PLUGIN_ROOT}/references/code-lookup.md). Report per-language: LSP backing reachable? For absent ones, suggest the standard install (pyright, gopls, typescript-language-server, rust-analyzer, etc.). Inform-only — never blocks. Record the per-language map in manifest.validation.lsp.servers (languages the user skips are omitted; the fallback ladder handles them).
• CLI probe. Walk the repo for tooling signals → likely CLIs (gh, aws, gcloud, az, kubectl, helm, terraform, docker, …) per ${CLAUDE_PLUGIN_ROOT}/references/cli-probe.md. which-check each candidate. Report two lists: available (found on PATH) and suggested (project signals indicate it'd be useful but it's missing — surface the install command from the reference). Record both in manifest.validation.clis.{available, suggested}. Inform-only — never blocks; the human installs. Don't suggest CLIs without project signals (no aws smell → don't ask about aws). The CLI probe runs before the environments interview because the available CLIs can help with that next step.
• Environments interview (all optional, batched in one AskUserQuestion): staging URL + deploy process; prod URL + deploy process + monitoring dashboards + alerting; feature-flag tool (launchdarkly | unleash | flagsmith | growthbook | other) + rollout policy. Record in manifest.validation.{staging, prod, feature_flags}. "Not yet" / "n/a" is a valid answer that omits the field. Where a relevant CLI from the probe is available, the agent should ask (consent-gated) whether to use it to discover or confirm env info — gh for deploy workflows, gcloud for GCP envs, kubectl for namespaces, etc. — rather than asking the human to type everything.
• Remote-agent readiness. Probe whether the project can delegate checkpoints to cloud Claude agents (vs. local subagents only):
• Automatable signals: git remote get-url origin — is the repo on a clonable host (github.com / a connected GHES)? Scan the repo's .mcp.json for type: stdio servers — a project-wide cloud blocker (stdio MCPs don't exist in the cloud VM), so flag it. These mirror the per-feature dae_delegable.py gate.
• Human one-time setup (DAE cannot automate this — surface it as a checklist the human confirms once): connect GitHub to claude.ai (/web-setup); create a cloud environment at claude.ai/code. Optional: install the Claude GitHub App on the repo (for PR/webhook triggers); create a routine (for RemoteTrigger-fired async/scheduled work — there is no create API, only the claude.ai UI).
• Record manifest.remote: ready: <true|false> (human confirms the one-time setup is done), channel: <isolation-remote | routine | none> (primary cloud vehicle; default isolation-remote — it needs no pre-created routine), stdio_mcp_blocker: <true|false>, and free-text notes. Inform-only — never blocks onboarding; when ready is false the dispatch router simply stays local (see references/handoff-dispatch.md).
• Autonomy proposal for Step 3's charter draft, based on what was found:
• staging + monitoring + feature flags → "high autonomy is well-supported by the validation infrastructure"
• staging + monitoring, no feature flags → "medium-to-high; pre-declare rollout paths in plans"
• staging only, no monitoring → "medium; add monitoring to expand the ceiling"
• none of the above → "low-to-medium until validation surface grows; recommend setting up staging + monitoring as an upstream backlog item"

The proposal is a recommendation — the human signs off (or overrides) at Step 3.

• Probe for project infrastructure. Before drafting the manifest, scan the repo for declarable infra dependencies and propose entries the human can accept/edit:

Present each draft entry to the human for confirmation, then add the approved entries to the manifest's infra: section per the schema (see engineer/references/handoff-dispatch.md + the schema in engineer/scripts/dae_resolve.py:_validate_infra).

Discovery is best-effort. If a project's infra doesn't fit the patterns above, ask the human to declare it manually — the declaration discipline is what makes downstream skills reliable.

• Capture infra quirks. After the infra entries are confirmed, batch one AskUserQuestion covering project-level runtime quirks that downstream skills need to know but can't discover from files:

Write the answers into manifest.infra_quirks — a project-level block consulted by engineer:fix Step 4, atdd:atdd test runs, and engineer:onboard gap-check. dae_resolve.py validates schema. Quirks are advisory metadata, not health probes — they exist so the next agent doesn't rediscover the same friction.

1. Draft the charter, get sign-off — draft CHARTER.md's 7 mandatory sections (methodology, architecture, conventions, scope, agent team, quality stance, autonomy stance). For an existing codebase, pre-fill what's inferable from the repo. Then present it and get the human's explicit confirmation — section by section for the judgment-heavy ones (scope, quality stance, autonomy stance + path overrides). Do not proceed to Step 4 until the charter is signed off.
2. Create the manifest — fill .engineer/manifest.yml (paths, roadmap/tracker, team, repos, quality thresholds, mutation, verification, autonomy, remote, agentic_summary).
3. Tracking decision — this is a human decision, not an agent default. Surface what the project appears to use (e.g. a repo full of Notion links → Notion) and ask the human to choose: notion | github-projects | linear | jira | local. notion: requires a connected Notion MCP — use it to create the tracker database (the TrackedFeature schema) or validate an existing one; DAE stores no API key (the MCP owns auth). local: feature folders are the tracker. Others: reserved — emit "not yet implemented". Never silently default to local to keep things moving. See references/tracker.md.
4. Bootstrap layout — create features/, empty .engineer/discussions.log; ensure .build/ is gitignored.
5. Discover features — walk the repo (read-only) for every feature-shaped chunk, documented and undocumented:

• Documented — Speckit specs/NNN-slug/, feature branches, docs/specs/*.md, GitHub Issues used as specs, informal README specs.
• Undocumented — feature-shaped code with no spec at all: scan the packages/modules for coherent capabilities (a route group, a service, a UI surface) that no document covers.

For each, record: source (or "code-only"), slug, state (spec-only / in-progress / shipped / merged), code co-locations (which packages/dirs the code lives in), and current DAE coverage (which of feature.md / acs.md / spec.md / acceptance tests exist — usually none). Greenfield project → discovery is empty; skip to Step 11.

1. Triage — with the human, rank the discovered features by importance to the project, and assign each a status (done shipped / in-progress / ready spec-only / parked dormant). Triage order drives the consolidation backlog's priority and which features get formalized first. Importance is a human judgement — surface a proposed ranking, let the human reorder.
2. Write the consolidation backlog + seed the tracker — two views of the triaged inventory:

• .engineer/consolidation.md: the inventory as a coverage table (one row per feature, a column per coverage artifact) plus consolidation tasks in triage-priority order. Goal stated at the top: every row all-✅. Each task — "bring feature X to full ATDD coverage" — is bounded and dispatchable to a remote agent; note the suggested execution mode per task.
• Seed the tracker — upsert a TrackedFeature row for every discovered feature (driver per references/tracker.md), not just the formalized ones, so the tracker shows the whole consolidation effort at a glance from day one. status from triage; checkpoint blank for features not yet in the DAE pipeline (consolidation.md tracks their coverage until they enter it).

1. Formalize the starting features — with the human, pick the 1–2 highest-triage features and run feature-init (onboarding-intake mode) on each now, so the project leaves onboarding with momentum. The rest stay as backlog tasks — do NOT formalize all of them in one onboard run.
2. Handoff — emit a summary; recommended_next points at the top consolidation-backlog task.

Migration is not done inside `onboard`. Moving specs/NNN-slug/ → features/NNN-slug/, backfilling acs.md/spec.md, and generating acceptance tests are consolidation tasks — worked down feature by feature after onboarding, via the pipeline (feature-init → discover-acs reverse-engineer mode → atdd:atdd → pipeline generation), and dispatchable to remote agents. onboard only discovers, triages, and plans.

Gap-check mode

Manifest exists → don't re-onboard. Validate: CHARTER.md has all 7 sections; manifest.yml schema-valid; features/ numbering monotonic; tracker config resolves; charter roles == manifest.team.default_roles. Report gaps with suggested fixes (mirrors consistency-check --project).

Strictly read-only. Forbidden:

• git checkout, git pull, git fetch, git merge, git rebase, git branch -d/-D — no git state mutations of any kind.
• git stash — even apparently-safe stashing changes working-tree state.
• Writing, editing, or deleting any file outside /tmp (no charter edits, no .engineer/ writes, no feature-folder touches).

If gap-check would benefit from a sync (e.g. the user is on a feature branch with merged remote), surface the suggestion in the report and stop. The caller invokes /engineer.post-merge or the manual git commands themselves. Do not infer "the user wants a sync" from "the user typed /engineer.onboard".

Full-onboard precondition

Full-onboard mutates the repo: writes CHARTER.md, .engineer/, features/. It must start from a clean state:

• HEAD on main/master (or the repo's default branch — read from git symbolic-ref refs/remotes/origin/HEAD).
• Clean working tree (git status --porcelain empty).
• Either: no current branch is a feature branch with unpushed work, or the user explicitly confirms "onboard on this branch".

If those conditions don't hold, stop and surface the situation — never auto-checkout, never auto-stash. Suggested wording: "Onboarding writes new files into the repo root. You're on <branch> with <N> unpushed commits — confirm you want to onboard here, or switch to main yourself first."

Handoff

Emit per ${CLAUDE_PLUGIN_ROOT}/references/handoff-summary.md. onboard is project-scope — its handoff goes to .engineer/handoffs/ (no feature folder exists). checkpoint: 0; artifacts: CHARTER.md, .engineer/manifest.yml. recommended_next: "per feature, /engineer.prime-context then /engineer.discover-acs; new ideas, /engineer.discuss".

If onboarding stopped because the human wasn't available to sign off the charter or choose the tracker, emit status: interrupted with human_action_needed: decision — naming exactly which decisions are outstanding.

References

• Foundation Design — charter format (§3), manifest schema (§2), storage layout (§1)
• Discuss & Upstream Funnel — methodology_root, onboarding intake
• references/tracker.md — the tracker drivers, setup, the Notion mapping
• Tracker Integration — the full contract
• engineer/scripts/dae_infra.py — what the manifest infra entries feed