j-mauro-r/SAT-Dengue-V2 / .agents/skills/deepworkplan
deepworkplan
VerifiedDeepWorkPlan — turn any repo AI-first and run Deep Work Plans. Routes to create, execute, refine, resume, status, verify, upgrade, and repo-onboarding sub-skills based on intent. Use when the developer wants to plan, execute, manage, or verify structured multi-task work, or make a repository AI-agent-ready.
guide/large-repo-onboarding.md
View on GitHub# DeepWorkPlan Guide — Onboarding a Large Repo as Its Own Deep Work Plan
> Part of the DeepWorkPlan methodology guide. The routing index is
> [`GUIDE.md`](GUIDE.md); read only the guide files your flow names.
## 15. Onboarding a Large Repo as Its Own Deep Work Plan
Making a repository AI-first (the `onboard` sub-skill) means analyzing the
**whole** codebase and **documenting all of it**: every `docs/` category, a
per-module `README.md` for each major module, and the full `.agents/` kit
(agents, skills, commands, catalogs). For a small or medium repo, the `onboard`
flow does this **inline** in one session — fast and correct.
For a **large** repo, doing it all inline is the wrong tool: it strains a single
context window, gives no per-artifact validation gate, can't be audited
task-by-task, and loses all progress if the session is interrupted. That is
exactly what Deep Work Plans exist to solve. So at scale, **onboarding becomes
its own Deep Work Plan** — the repo's first plan is literally "finish onboarding
myself," executed task-by-task with gates and full resumability. The methodology
dogfoods itself: onboarding uses the very loop it installs.
**When the `onboard` flow chooses the plan-driven path** (its Phase 2b decision):
more than ~8 major modules, a monorepo / multi-package workspace, an orchestrator
hub, ~15+ total artifacts to generate, or the developer asks. Otherwise it stays
inline (the typical case).
**Shape of the onboarding plan** (one task = one artifact, each with a gate):
- **Task 1** — `AGENTS.md` + `CLAUDE.md` up front, so every later task has the
index and mandatory rules to anchor to.
- **One task per `docs/` category** — each gated on "real repo-specific content,
no placeholders, real commands, links resolve."
- **One task per major module** — its `README.md` (responsibility, public
surface, how it's tested). This is the part that scales worst inline.
- **One task for the `.agents/` kit** — agents + skills + commands + catalogs,
gated on "catalog matches disk."
- **One task to install the skill + scaffold `.dwp/` and `tmp/`** and offer addons.
- **Mandatory final task** — the onboard Phase 8 conformance self-check (run
`/dwp-verify`), folded into the spec-mandatory **Final Review** (security pass,
final-state validation, skills reconciliation, documentation reconciliation).
The Executive Report is optional.
Each task reuses the **repo's real validation command** as its gate (lint,
`md`-check, link-check, tests). The plan is created with `/dwp-create`, refined
with `/dwp-refine`, executed with `/dwp-execute`, resumed with `/dwp-resume`, and
inspected with `/dwp-status` — the same loop as any other plan. The full task
shapes live in the onboard sub-skill's
`onboard/templates/onboarding-plan.md` reasoning aid.
> **The Phase 0 onboarding rules do not relax.** Non-destructive, idempotent, and
> **no-placeholders** move *into* each task's Acceptance Criteria and gate; they
> do not disappear. A plan-driven onboarding that ships a placeholder doc is the
> same failure as an inline one.
This is distinct from the **orchestrator plan** (§13): orchestrator plans
coordinate child DWPs across **multiple repos**; the onboarding plan completes
documentation work within **one** large repo. A large monorepo may use both — an
onboarding plan to document itself, then orchestrator plans for cross-package
features.
---
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |