Przejdź do treści
Biblioteka skilli
Programowanie · Planowanie

planning

/planning

Planuje implementację oprogramowania z opcjonalnym zapisem w plikach: trwały zapis nierozstrzygniętych decyzji o architekturze i zakresie oraz stan odtworzenia dla wielofazowych prac.

Repozytorium
ai-skills ↗
Licencja
MIT
Ostatnia zmiana repozytorium
25.10.2026

Ocena i bezpieczeństwo

100/ 100Klasa: dobry
Format25 / 25
Frontmatter z nazwą i opisem, poprawna nazwa, opis do 1024 znaków, nazwa zgodna z folderem.
Odkrywalność25 / 25
Opis mówi, kiedy użyć skilla, ma sensowną długość i jest w 3. osobie.
Treść35 / 35
Co najmniej 150 słów i do 500 linii, nagłówki, listy kroków, przykłady albo kod.
Higiena15 / 15
Bez zaślepek typu TODO, bez ścieżek Windows, z licencją.
Ryzykobrak sygnałów w skanie
Pliki w paczce
10
Skrypty
1
Komentarze HTML
0
Adresy URL w SKILL.md
0
Przestarzałość

Brak sygnałów. Repozytorium zmienione 25.10.2026, bez odwołań do wycofanych modeli i API.

Licencja

MIT, otwarta: można używać i zmieniać, z zachowaniem informacji o autorze.

Co sprawdzić przed instalacją
  • 1 skrypt w paczce. Przejrzyj je przed instalacją, bo Claude może je uruchomić.

Skrypty: scripts/init-plan.sh

Co robi i kiedy

Planuje implementację oprogramowania z opcjonalnym zapisem w plikach: trwały zapis nierozstrzygniętych decyzji o architekturze i zakresie oraz stan odtworzenia dla wielofazowych prac.

Oryginalny opis (EN)

Software implementation planning with optional file-based persistence. Use when asked to plan, when unresolved architecture or scope decisions need a durable record, or when multi-phase implementation needs recovery state. For the full research-and-issue workflow, use the ia-plan command (/ia-plan in Claude Code).

Jak działa

  1. Procedure
  2. Durable state
  3. Conditional guidance
  4. Authority and verification
Sekcje z nagłówków SKILL.md, w oryginale · 39 linii, 689 słów
Podgląd SKILL.md
---
name: planning
class: workflow
description: >-
  Software implementation planning with optional file-based persistence. Use
  when asked to plan, when unresolved architecture or scope decisions need a
  durable record, or when multi-phase implementation needs recovery state.
  For the full research-and-issue workflow, use the ia-plan command (/ia-plan
  in Claude Code).
---

# Planning

Produce the smallest plan that reduces implementation risk and preserves state costly to reconstruct. User requirements and verified facts govern the plan; a reference implementation supplies evidence about behavior, not independent authority to change scope.

## Procedure

1. Define the concrete outcome, evidence, binary or quantitative success threshold, scope boundaries, and stop conditions. If the request already names an artifact and success signal, use them. Otherwise repair activity-only goals before planning.
2. Separate the objective from the proposed mechanism: would the goal remain valid if the implementation changed? Ensure a colleague can understand the objective alone and verify it without knowing component internals.
3. Choose a full durable plan for interdependent phases, recovery needs, or work crossing context limits; an inline list for clear work fitting one session; direct implementation for clear scope and acceptance criteria. File and tool-call counts are signals, not triggers. Inspect hidden decisions such as cache TTL, invalidation, and key shape before treating work as simple.
4. Inspect existing tests, canonical commands, related code, and any existing brainstorm specification. Extend repository patterns. Record significant architectural tradeoffs in its ADR convention when needed.
5. Define files, ownership, interface contracts, dependencies, concrete tasks, verification, and exit conditions. Organize phases around runnable capabilities with implementation and tests together. Put high-variance decisions before mechanical work; keep execution dependencies in phase order.
6. Verify the plan against the user's requirements. Keep phase status and next step consistent. Never weaken acceptance criteria to match incomplete code or add speculative checks and machinery.
7. Deliver the plan when planning alone was requested. Continue implementation when already authorized. Ask about execution mode only if it materially changes cost, risk, isolation, or review quality.

## Durable state

For a full working plan, read [plan-format.md](./references/plan-format.md) and scaffold with [init-plan.sh](./scripts/init-plan.sh), anchored to the installed skill directory rather than the caller's working directory. When the project already uses a spec or plan system, keep that system's artifact format; the skill owns the clarification, content, and approval gates, not the representation. Otherwise, use `.plan/task_plan.md` for uncommitted session state and `docs/plans/` for a formal committed plan.

Inspect an existing plan before overwriting it. Continue the same work in place; ask which plan wins when different work would displace unchecked tasks. Never silently discard or bulk-close open work, including tracker items. Use `--force` only after resolving that choice. Add secondary artifacts only for requested deliverables or state the main plan cannot express clearly.

## Conditional guidance

- For vague goals, a proposed mechanism disguised as an objective, or choosing planning depth, read [goal-definition.md](./references/goal-definition.md).
- For a full plan, template, test discovery, source pointers, interface choices, or plan-quality review, read [plan-format.md](./references/plan-format.md). Preserve exact specification values and name concrete paths or commands in every task.
- When decomposing phases, setting ownership, or deciding what needs clarification, read [task-design.md](./references/task-design.md). Respect existing user choices and project constraints before selecting technical details. Escalate materially different outcomes or authority gaps; routine implementation choices stay with the implementer.
- For vertical slicing, checkpoints, reference semantics, phase posture, or delegated execution, read [execution-and-methodology.md](./references/execution-and-methodology.md).
- When starting multi-phase work or resuming after a gap, read [operational-patterns.md](./references/operational-patterns.md) for recovery, context checks, and error handling.
- When deepening an existing plan, read [plan-deepening.md](./references/plan-deepening.md); preserve its structure and add targeted research.
- When selecting execution posture, handing off, or integrating related skills, read [execution-handoff.md](./references/execution-handoff.md).

## Authority and verification

Use the active harness's question tool when its schema and context permit; otherwise ask in chat. Batch material unknowns, make safe assumptions for routine details, and never manufacture a question count.

Verify that each task is verb-first, concrete, independently verifiable, and comprehensible without conversation history. Check naming and interface consistency across tasks. Every phase needs a runnable capability, explicit verification, and a context-safe recovery boundary; process items must name the capability or observed defect they gate. Open questions contain only blocking unknowns. A plan artifact does not itself establish implementation progress.

For ambiguous requirements use `ia-brainstorming`; for significant new trust boundaries, auth, payments, or external APIs, obtain the applicable security threat-model review before implementation. Use `ia-writing` for prose. Report the plan location, material decisions, and remaining blockers without claiming unexecuted checks passed.
Pliki w paczce (10)
  • SKILL.md 6 KB
  • SPEC.md 4 KB
  • references/execution-and-methodology.md 5 KB
  • references/execution-handoff.md 3 KB
  • references/goal-definition.md 5 KB
  • references/operational-patterns.md 4 KB
  • references/plan-deepening.md 2 KB
  • references/plan-format.md 8 KB
  • references/task-design.md 4 KB
  • scripts/init-plan.sh 2 KB

Wersje

RepozytoriumWynikLicencja
Ta strona · iliaal/ai-skills100MIT
nimbalyst/nimbalyst100MIT
open-metadata/openmetadata100Apache-2.0
davila7/claude-code-templates100MIT
smart-ai-memory/attune-ai100Apache-2.0
iliaal/whetstone100MIT
doloopz94/virtual-office-muse-ai100MIT
microclaw/microclaw100MIT
jkitchin/skillz100MIT

Migawka bazy z 13.11.2026 · wersja paczki 398d2987fe4f (skrót SHA-256 całej paczki)

Pobrania w ostatnich 30 dniach: 1 · ranking

Zgłoś problem z tym skillem: szkodliwa treść, błędny opis albo prośba autora o usunięcie.

Podobne skille

  • 90teamimplement
    Prowadzi wieloetapowe wdrożenie oprogramowania zespołem inżynierów: research, plan, implementacja, przegląd i walidacja QA według PLAN.md.
  • 100mine-plan-create
    Tworzy lub aktualizuje plan implementacji oprogramowania oparty na dowodach i architekturze, na tyle precyzyjny, by niezależni agenci kodujący wykonali go bez ponownego odkrywania decyzji projektowych.
  • 100plan
    Pisze zwalidowany plan wdrożenia w oparciu o kod, dla złożonych zmian: nowych funkcji, zmian architektury, pracy na wielu plikach i niejasnego zakresu.
  • 100crispi-planning
    Prowadzi planowanie złożonych zadań wieloplikowych (migracja, refaktoryzacja przekrojowa, zmiana narzędzi) w pięciu fazach CRISPI i zapisuje plan w .claude/plans/ między sesjami.
  • 100planning-android-implementation
    Projektuje architekturę i planuje fazy implementacji funkcji w Androidzie Bitwarden: inwentarz plików, podział na fazy i decyzje projektowe przed kodowaniem.
  • 100plan-mode-review
    Przegląda plany i pull requesty w pięciu fazach: architektura, jakość kodu, testy, wydajność i gotowość produkcyjna.

Uwagi i zgłoszenia

Wiadomość trafi do Bartosza Sokolińskiego. Adres e-mail jest potrzebny tylko do odpowiedzi. Szczegóły w polityce prywatności.