Przejdź do treści
Biblioteka skilli
Zarządzanie i produkt · Research

discover

/discover

Bada dostępne informacje, szukając niewiadomych i ślepych punktów, aby zbudować rzetelne rozumienie problemu przed działaniem.

Licencja
MIT
Ostatnia zmiana repozytorium
25.10.2026

Ocena i bezpieczeństwo

87/ 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.
Treść22 / 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
1
Skrypty
0
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 robi i kiedy

Bada dostępne informacje, szukając niewiadomych i ślepych punktów, aby zbudować rzetelne rozumienie problemu przed działaniem.

Oryginalny opis (EN)

Use when a question or problem is identified but what needs to be known about it must be established first — the unknowns themselves are not yet identified. Explore available information, surface blind spots, and form an evidence-based understanding the user can act on.

Jak działa

SKILL.md bez nagłówków drugiego poziomu · 15 linii, 550 słów

Podgląd SKILL.md
---
name: discover
description: Use when a question or problem is identified but what needs to be known about it must be established first — the unknowns themselves are not yet identified. Explore available information, surface blind spots, and form an evidence-based understanding the user can act on.
---

Starting from an identified question or problem, discover and connect the knowledge, context, and evidence needed to understand the subject, and form an understanding that lets a user unfamiliar with the subject judge and act on their own. Assume the user may not know the right search terms, where materials are, or even what they need to know. If the question or problem itself is not identified, state what is missing and stop.

Establish what the user currently knows, assumes, and already recognizes as unknown. Investigate available information autonomously — materials, code, data, history, whatever the environment provides — before asking the user. Ask only for what the user alone can provide, one question at a time, waiting for each answer. When needed knowledge is not at hand, trace available clues toward where it may exist; if the trail runs out, state what was checked, what is missing, and how the gap affects understanding.

Do not stop at the questions the user has raised. Critically examine the current understanding, assumptions, scope, and sources to surface important unknowns and blind spots not yet recognized: implicit premises, missing stakeholders or cases, exceptions and boundaries, effects on other work or systems, gaps between current behavior and intended specification, constraints that may no longer hold, or flaws in the premise of the question itself. Do not enumerate possibilities without limit; prioritize the unknowns that most affect the next judgment or action, and reduce them by whatever method fits — investigation, dialogue, or concrete examples.

Do not accept load-bearing premises unconditionally, whether they come from the user or from materials; consider whether they are unverified, biased, or wrong when they would sway the conclusion. Distinguish confirmed facts from agreed decisions, hypotheses, inferences, and unverified assumptions — including whether user-provided information is direct observation, judgment, or hearsay — so the reader cannot mistake one for another. Give load-bearing facts a source that can be re-verified; do not treat information without evidence as confirmed fact. When important sources conflict, surface the conflict rather than silently picking a winner; resolve it when the next step depends on it, and otherwise state what remains undecided and what it affects.

Connect what was found into a coherent explanation of the subject rather than a list of findings. Draw the conclusions the evidence supports; where evidence is insufficient or multiple interpretations remain, do not force a single conclusion — state the uncertainty and what additional information would resolve it.

Stop when the unknowns that matter for the next step are sufficiently reduced, the remaining unknowns, uncertainties, and risks are explicit, and the user can understand the subject and decide or act on their own — not when every unknown is resolved. If required information cannot be obtained, present what was confirmed, what is missing, and how that affects understanding, then stop. Present a self-contained synthesis that a person or agent without the conversation history can use to take over: the question, key background, confirmed facts with their evidence, the understanding formed, important premises, discovered blind spots, remaining unknowns and their impact, and considerations for the next decision or action.

Do not decide what to work on, choose which hypothesis to test first, produce an implementation specification, carry out the solution, change production code or environments, or document the whole domain exhaustively.

Wersje

RepozytoriumWynikLicencja
Ta strona · 70-10/claude-code-plugins87MIT
DiogoHSM/driver100MIT
taishi-i/awesome-japanese-nlp-resources100CC0-1.0
SergioLizana/Dynamo100MIT
eliyce/paqad-ai100MIT
hunter-io/claude-plugin100MIT
konglong87/superpm100MIT
joestump/claude-plugin-sdd100MIT
ArthurgLeonida/paperscout100MIT

Migawka bazy z 13.11.2026 · wersja paczki e8267b911296 (skrót SHA-256 całej paczki)

Pobrania w ostatnich 30 dniach: 0 · ranking

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

Podobne skille

  • 100explore-unknowns
    Mapuje niejasne zadanie przed planem: wyjaśnia niejasności, nieznany kod i odchylenia od planu, by zbudować wspólne rozumienie.
  • 90explore-unknowns
    Prowadzi krok po kroku przez mapowanie niewiadomych zadania w czterech kwadrantach, zaczynając od tego, co już wiadomo.
  • 90explore-unknowns
    Prowadzi przez kwadrantową mapę niewiadomych zadania: znane znane, znane nieznane, nieznane znane i nieznane nieznane, gdy prośba jest niejasna lub dziedzina obca.
  • 100025-quality-attribute-discovery
    Identyfikuje i priorytetyzuje atrybuty jakości, czyli wymagania niefunkcjonalne, problemu przed architekturą i projektem.
  • 90recipe-discover
    Ujmuje szanse produktowe i tworzy istotne dla decyzji hipotezy z dostępnych dowodów, gdy bada się problem, szansę rynkową lub potrzebę użytkownika.

Uwagi i zgłoszenia

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