Skip to content

tibursocampos/agent-dev-toolkit / core/skills/sdd-plan

sdd-plan

Verified

Create a baby-step PLAN from an existing PRD (agent PLAN .md; language per LANGUAGE.md + preferences). Feeds sdd-develop. Use when creating a plan or invoking /sdd-plan.

references/command.md

View on GitHub
# sdd-plan — command playbook

**Load after Step -1 gates.** Ordered step discovery without dumping `SKILL.md` Process. Load **one** section file per step (`SKILL-REFERENCE-RETRIEVAL.md`).

| Step | Action | Lazy section / contract |
|------|--------|-------------------------|
| -1b | Caveman Lite when active | `CAVEMAN.md` |
| -1 | Pipeline + STORAGE; `invocation_context` | `PIPELINE.md`, `STORAGE.md`, `INVOCATION-CONTEXTS.md` |
| 0 | Workspace; stack; classic feature root | `STORAGE.md` |
| 1 | Resolve PRD (canonical path only) | `references/filename-numbering.md` |
| 2–4 | Resolve `direct` vs `orchestrated`: direct derives PLAN-native steps from the approved PRD and never acquires O1 checklist inputs; O2 validates/acquires its required checklist through its owner, then preserves those IDs. Two self-reviews before preview | `references/checklist-acquisition.md`, `references/baby-step-sizing.md`, `references/self-review.md`, `references/challenge-vagueness.md`, `references/plan-magro.md`; `LIVE-STAGE-TABLE.md` § sdd-plan |
| 5.5 | PLAN storage path | `references/storage-gitignore.md` |
| 5.75 | Confirm before write (**sim**) | `PIPELINE.md` |
| 6 | Write PLAN (Agent + sim; magro; Execution policy; Mapa REQ) | `references/template-usage.md`, `references/plan-magro.md` |
| 6.5 | Structural validate (`validate-plan`) | `references/validate-plan.md` |
| 7 | Validate with user; handoff develop | `references/quality-checklist.md`, `references/status-legend.md` |

## Stage table

Standard, state keys, glyphs, and label map: `LIVE-STAGE-TABLE.md` (§ sdd-plan). Columns, in order: `stage`, `status`, `weight`, `what_happened`. Headers and state labels follow the user chat language; glyphs are the ASCII ones from the standard.

| Stage | Weight | SKILL step |
|-------|--------|------------|
| Resolve PRD | Medium | 1 |
| Explore the repo | Very high | 2-4 item 5 (repository evidence) |
| Draft | High | 2-4 items 1-3 (derive steps, REQ map) |
| Review 1 | High | 2-4 item 4, first pass |
| Review 2 | High | 2-4 item 4, second pass |
| Content gate | Low | 5.75 |
| Preview | Low | 5.9 (summary before the write question) |
| Write gate | Low | 5.9 (write confirmation) |
| Persist | Low | 6, 6.5 |

You must render the full table at each of these points:

1. When the flow starts: first row `in_progress`, the rest `pending`.
2. Before each gate (each question to the operator), including the content gate and the write gate.
3. After each gate answer.
4. On each row state change, including the start and end of Review 1 and Review 2.
5. On a blocker or failure (a failed second review, an open question, a failed validation): stop the redraw at `blocked`, name the cause, and do not present a successful table.
6. At close or handoff.

Completeness: always render the full table with every row. Do not summarize rows, show only the changed row, or omit completed rows. `skipped` requires a reason in `what_happened`.

Gate answers use **sim** / **ajustar** / **cancelar**. Do not save this table in the PLAN.

Selective retrieval: `references/selective-retrieval.md`. Product docs language: `references/product-docs-language.md`.
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: 468kUpdated Claude CodeCodexVerifiedInstall
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: 321kUpdated Claude CodeCodexVerifiedInstall
dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependenciesStars: 321kUpdated Claude CodeCodexVerifiedInstall
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 availableStars: 321kUpdated Claude CodeCodexVerifiedInstall