Przejdź do treści
Biblioteka skilli
Programowanie · Research

opensourcefaq

/opensourcefaq

Odpowiada na pytania o produkty open source na podstawie kodu, dokumentacji, DeepWiki i aktualnych źródeł, zapisując odpowiedź z przypisami.

Repozytorium
blog ↗
Licencja
LGPL
Ostatnia zmiana repozytorium
21.10.2026

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
2
Skrypty
0
Komentarze HTML
0
Adresy URL w SKILL.md
0
Przestarzałość

Brak sygnałów. Repozytorium zmienione 21.10.2026, bez odwołań do wycofanych modeli i API.

Licencja

LGPL, otwarta.

Co robi i kiedy

Odpowiada na pytania o produkty open source na podstawie kodu, dokumentacji, DeepWiki i aktualnych źródeł, zapisując odpowiedź z przypisami.

Oryginalny opis (EN)

Answer open-source product questions by combining local source-code and documentation research, project framework notes generated by init, DeepWiki repository knowledge, and current web evidence, then save a sourced, diagram-rich Markdown answer to the current project's markdown directory. Use when the user asks about behavior, architecture, internals, configuration, APIs, troubleshooting, performance, operations, extension points, practical reproduction, or version-specific details of one or more open-source projects and provides the question, all relevant local source directories, and DeepWi

Jak działa

  1. Overview
  2. Inputs
  3. Workflow
  4. Practical Reproduction Requirements
  5. Article Structure
  6. Validation Rules
Sekcje z nagłówków SKILL.md, w oryginale · 130 linii, 1 111 słów
Podgląd SKILL.md
---
name: opensourcefaq
description: Answer open-source product questions by combining local source-code and documentation research, project framework notes generated by init, DeepWiki repository knowledge, and current web evidence, then save a sourced, diagram-rich Markdown answer to the current project's markdown directory. Use when the user asks about behavior, architecture, internals, configuration, APIs, troubleshooting, performance, operations, extension points, practical reproduction, or version-specific details of one or more open-source projects and provides the question, all relevant local source directories, and DeepWiki repoName values.
---

# Opensourcefaq

## Overview

Use this skill to produce a verified, source-backed, diagram-rich Markdown answer for questions about open-source products. Require the user to provide the question, every relevant project source directory, and the corresponding DeepWiki repoName values; remind them to clone the source first and run `init` to generate project framework notes before using the skill.

## Inputs

Confirm these inputs before writing the final article:

- `question`: the exact technical question to answer.
- `source_dirs`: absolute paths to all open-source project checkouts involved.
- `deepwiki_repo_names`: DeepWiki repoName values such as `owner/repo` for each relevant project.
- `version_context`: branch, tag, commit, release, or product version if the answer depends on version-specific behavior.

If any required input is missing, ask for it. If only the version context is missing, inspect the local checkout with `git -C <dir> rev-parse --abbrev-ref HEAD`, `git -C <dir> rev-parse HEAD`, and release files when useful, then state the inferred version and its limits.

## Workflow

1. Restate the question and scope.
   - Identify the involved projects, versions, modules, configuration, and runtime assumptions.
   - Split broad questions into concrete subquestions that can be checked against source code.

2. Analyze the problem.
   - Form hypotheses from the user's wording, known architecture, and likely code paths.
   - Name ambiguity early instead of silently choosing one interpretation.
   - Prefer the simplest explanation that source code and docs support.

3. Prepare evidence.
   - Search local source directories first with `rg`, `rg --files`, and targeted file reads.
   - Search both implementation code and project documentation relevant to the question. Include source files, tests, examples, SQL/API definitions, configuration templates, README/manual/reference docs, release notes, design docs, and migration notes when present.
   - When code and docs disagree, verify against the code path and note the documentation drift.
   - Read the project framework notes generated by `init` when present, then verify important claims against source files.
   - Browse DeepWiki project overviews for the listed repoName values.
   - Ask DeepWiki focused questions about architecture, relevant modules, and the exact technical behavior under investigation.
   - Search the web for relevant articles, issues, release notes, documentation, design docs, and discussions when current or external context matters.

4. Filter evidence quality.
   - Prefer source code, official docs, release notes, design docs, and maintainer comments.
   - Treat blogs, forum answers, and copied snippets as secondary evidence.
   - Check publication date, project version, branch, API names, file paths, and whether the behavior still exists in the local checkout.
   - Discard or clearly qualify outdated material.

5. Write the Markdown article.
   - Save under the current project's `markdown/` directory. Create that directory if it does not exist.
   - Use a clear filename derived from the question, for example `markdown/opensourcefaq-<topic>.md`.
   - Cite local files with paths and line numbers when possible.
   - Cite DeepWiki and web sources with links or repoName references.
   - Make the article diagram-rich. Include Mermaid diagrams for architecture, flow, sequence, state, dependency, data model, or query plans whenever they clarify the answer.
   - Add tables for evidence, version differences, tradeoffs, parameters, or operational checklists when useful.
   - If the user asks for practical execution, troubleshooting, benchmarking, reproduction, or "how to use", include a complete simulated environment and sample data generation method unless the real environment is already provided and sufficient.
   - Keep the article focused on the user's question; do not expand into an unrelated product overview.

6. Verify before finalizing.
   - Re-check every important technical claim against local code and DeepWiki.
   - Verify commands, configuration examples, API names, SQL, YAML, code snippets, and file paths.
   - If a claim cannot be verified, mark it as an inference and explain the basis.
   - If code behavior differs across versions, state exactly which version was verified and what may differ elsewhere.
   - For practical examples, verify that setup steps, generated data shape, commands, and expected outputs are internally consistent. If commands were not actually run, say so explicitly.

## Practical Reproduction Requirements

When the answer includes hands-on usage, troubleshooting, benchmarking, or an executable example, include:

- A minimal simulated environment: container commands, local build steps, package prerequisites, service configuration, database/schema setup, or equivalent project-native setup.
- Data generation: SQL, scripts, fixtures, API calls, or command sequences that create representative data covering normal cases and edge cases.
- Execution steps: commands or queries the reader can run in order.
- Expected observations: key outputs, metrics, query plans, logs, errors, or screenshots/diagrams to compare against.
- Cleanup steps when the setup creates services, containers, databases, files, or test data.

Prefer project-native tooling already present in the repository. Do not invent heavyweight infrastructure when a lightweight local or containerized setup demonstrates the point.

## Article Structure

Use this structure unless the user's question calls for a shorter answer:

```markdown
# <Question as a precise title>

## 问题描述
<Restate the question, scope, projects, and verified versions.>

## 结论先行
<Direct answer in a few paragraphs.>

## 问题分析
<Break down the reasoning path and key assumptions.>

## 准备素材
### 本地代码与文档搜索
<Important implementation files, tests, examples, docs, release notes, configs, functions, modules, and line references. State any mismatch between code and docs.>

### DeepWiki 综述与问答
<Relevant architecture summary and focused DeepWiki answers.>

### 网络资料筛选
<Current, high-quality external sources and rejected outdated sources when relevant.>

## 原理与实现细节
<Explain mechanisms, data flow, call flow, configuration, and edge cases.>

## 图解
<Mermaid architecture/flow/sequence/data-model diagrams and, when useful, tables that make the answer visually clear.>

## 实操示例
<If practical execution is relevant, provide simulated environment setup, sample data generation, executable commands, expected observations, and cleanup. Otherwise explain why no hands-on reproduction is needed.>

## 拓展思考
<How to generalize the method, compare adjacent designs, or avoid common mistakes.>

## 验证记录
<What was checked in source code, DeepWiki, tests, or commands, and any remaining uncertainty.>

## 参考资料
<Local paths, DeepWiki repoNames, and web URLs.>
```

## Validation Rules

- Do not rely on memory alone for product behavior.
- Do not cite web articles without checking version freshness and consistency with local source.
- Do not treat DeepWiki as final authority when local code contradicts it; source code wins.
- Do not invent line numbers, functions, flags, or configuration names.
- Do not omit related local documentation when searching source directories; code-only evidence is insufficient unless the project has no relevant docs.
- Do not output a text-only article when diagrams or tables would clarify the answer.
- Do not provide practical SQL/code/CLI examples without a reproducible environment and sample data plan, unless the user explicitly asks for conceptual explanation only.
- Keep final Markdown in Chinese by default unless the user asks for another language.
- End the response to the user with the saved file path and a concise verification summary.
Pliki w paczce (2)
  • SKILL.md 8 KB
  • agents/openai.yaml 1 KB

Wersje

RepozytoriumWynikLicencja
Ta strona · digoal/blog100LGPL
digoal/blog100LGPL

Migawka bazy z 13.11.2026 · wersja paczki 7b6458786625 (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

  • 100pgfaq
    Odpowiada na pytania o rdzeń PostgreSQL na podstawie lokalnych źródeł, dokumentacji w doc/src/sgml i DeepWiki, zapisując opisaną odpowiedź w Markdown.
  • 100mcp-deepwiki
    Prowadzi przez użycie DeepWiki MCP do przeglądania dokumentacji repozytoriów GitHub: struktura repozytorium, API bibliotek i pytania o projekty open source.
  • 100deepwiki-api
    Przegląda dokumentację repozytoriów GitHub przez API DeepWiki wywoływane curlem: struktura repozytorium, API bibliotek i pytania o projekty open source, bez serwera MCP.
  • 100open-source-project-article
    Analizuje projekt open source z adresu repozytorium i pisze obszernie źródłowany artykuł po chińsku w Markdown: README, architektura z DeepWiki MCP, konkurenci i weryfikacja twierdzeń.
  • 100researching-with-deepwiki
    Bada repozytoria GitHub, GitLab i Bitbucket przez serwer DeepWiki, odpowiadając na pytania o architekturę obcych projektów open source.
  • 100deepwiki-cli
    Odpowiada na pytania o architekturę, API i wnętrze repozytorium GitHub, którego nie ma lokalnie, oraz o działanie biblioteki lub frameworka, przez deepwiki-cli.

Uwagi i zgłoszenia

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