Dev & Engineering ✓ Microsoft · Official repository-analysissource-code-analysiscode-navigationgitlinked-citations

Repository Wiki Q&A

Answers codebase questions by tracing behavior directly to source files.

FollowSkills review · FSRS-2.0
Use with care
52/ 100 5-point scale 2.6 / 5
1 2 3 4 5 6
1Trust16 / 25 · 3.2/5

The skill restricts answers to source-file analysis and requires citations; no malware, credential theft, covert exfiltration, or destructive default is evident, supporting a moderate trust score. Deductions apply because it mandates git remote/branch commands and user repository disclosure without explaining sensitive-repository handling, permission boundaries, failure recovery, or data flows.

2Reliability8 / 20 · 2.0/5

The main procedure is coherent: resolve context, search files, read evidence, and synthesize an answer. Deductions apply because it gives no diagnostic path for command failure, missing remotes, detached HEAD, inaccessible repositories, or citation-generation failure; static review also lacks executable evidence specific to this skill, so the score remains at or below the static ceiling of 10.

3Adaptability10 / 15 · 3.3/5

Activation conditions are clear and include repository behavior, files, functions, and codebase navigation; it also requests the user's language. Deductions apply because non-fit ranges, input/output boundaries, and handling across hosting platforms are unspecified; remote citations depend on external GitHub/Azure DevOps URLs, with little discussion of mainland-China reachability or constrained-network use.

4Convention8 / 15 · 2.7/5

The SKILL.md has a name, description, MIT license, author, version, and readable sections. Deductions apply because it lacks examples, FAQs, troubleshooting, known limitations, a changelog, and explicit skill-level maintenance ownership or update path; its output requirements are heavy and do not explain progressive disclosure for complex answers.

5Effectiveness6 / 15 · 2.0/5

The core goal—answering repository questions from source evidence—is clear, and citations plus a key-files table could produce usable results. Deductions apply because mandatory Mermaid diagrams, tables, and specific Markdown formatting are unsuitable for every question, representative outputs are absent, and unresolved context can block otherwise answerable tasks; under static calibration the score cannot exceed the near-7 range.

6Verifiability4 / 10 · 2.0/5

The source-only rule and inline citation requirement provide some auditability, and the repository includes general CI and test material. Deductions apply because the supplied tests do not cover wiki-qa key paths and there is no skill-specific evaluation, independent review, or execution evidence; static evidence therefore supports only a limited score.

Evidence confidence:Low Reviewed Jul 20, 2026 Reviewed revision 9ccaf7c3704a
Before you use it
  • The skill blocks answering until git remote and branch context are resolved, but does not define fallbacks for no remote, detached HEAD, command failure, or unreachable networks.
  • Mandatory remote links, Mermaid diagrams, and tables may increase friction for Chinese users and constrained-network environments.
  • Repository-level CI and tests do not establish correctness or citation reliability for this specific wiki-qa skill.
See the full review method →

What does this skill do, and when should you use it?

Repository Wiki Q&A is a software-development Agent Skill for understanding how an existing codebase works. It resolves the repository context first, then searches and reads relevant source files before composing an evidence-based answer. Responses follow the question’s language and include line-level citations. It fits developers navigating unfamiliar repositories, locating definitions, or explaining implementation behavior.

Runs git remote get-url origin to inspect the repository remote and git rev-parse --abbrev-ref HEAD to determine the branch; asks whether the repository is local-only or has a source URL; searches and reads relevant source files; synthesizes the findings into an answer with linked remote citations or local file-and-line citations; and identifies missing evidence when the available files are insufficient.

  1. A developer joining an unfamiliar project wants to understand how a component works.
  2. A maintainer needs to find where a particular file, function, or behavior is defined.
  3. A teammate asks how a feature works and needs an answer grounded in implementation evidence.
  4. A technical writer needs a source-cited repository explanation with key files and, where relevant, an architecture diagram.

What are this skill's strengths and limitations?

Pros
  • Requires repository remote and branch context to be resolved first.
  • Explicitly grounds answers in source-code evidence.
  • Supports linked remote citations or local path-and-line citations.
  • Targets practical code understanding, definition lookup, and repository navigation.
Limitations
  • Requires the agent to run Git commands and read repository files.
  • The user must resolve the repository context before the analysis proceeds.
  • The SKILL.md does not provide a test suite or specify supported programming languages.
  • The source does not document behavior for inaccessible remotes or unavailable source files.

How do you install this skill?

Install the collection with npx skills add microsoft/skills, then select the required skill in the wizard. In Copilot CLI, you can also run /plugin marketplace add microsoft/skills, followed by /plugin install deep-wiki@skills. The source material does not document a standalone installation command for .github/plugins/deep-wiki/skills/wiki-qa/SKILL.md.

How do you use this skill?

After installation, ask a concrete repository question such as: “How does authentication work in this codebase, and where is it defined?” Before answering, provide a source repository URL or confirm that the repository is local-only.

FAQ

Is this intended for general programming questions without a repository?
No. Its activation and procedure focus on repository, source-file, function, and component analysis.
Do I need to provide a GitHub URL?
You must resolve the repository context first: provide a source repository URL for a remote repository or confirm that the repository is local-only.
What evidence does it cite?
Remote repositories use branch- and line-linked citations; local-only repositories use file paths and line numbers.
Does it require an MCP server?
The SKILL.md does not declare an MCP server requirement.

More skills from this repository

All from microsoft/agent-skills

Dev & Engineering ✓ Microsoft · Official

Wiki Codebase Researcher

Trace real code paths across files and produce evidence-backed system analysis.

Writing & Content ✓ Microsoft · Official

Wiki Page Writer

Generates evidence-backed technical Wiki pages with source citations and deep Mermaid diagrams.

Dev & Engineering ✓ Microsoft · Official

Codebase Onboarding Guide Generator

Generate source-grounded onboarding documentation for contributors, engineers, leaders, and product managers.

Dev & Engineering ✓ Microsoft · Official

Wiki Architect

Turns a code repository into a structured documentation catalogue and guided onboarding path.

Dev & Engineering ✓ Microsoft · Official

AGENTS.md Context Builder

Generates actionable AGENTS.md guidance for repository folders that lack coding-agent context.

Dev & Engineering ✓ Microsoft · Official

Wiki Changelog Assistant

Turn Git history into a structured, categorized changelog.

Automation & Ops ✓ Microsoft · Official

Azure Reliability Advisor

Assess and improve reliability for Azure Functions and App Service.

Dev & Engineering ✓ Microsoft · Official

Azure AI Text Analytics for Python

Guides Python developers through Azure-powered text analysis and NLP workflows.

Dev & Engineering ✓ Microsoft · Official

Azure AI Vision Image Analysis

Guides Python coding agents in implementing Azure AI Vision image understanding.

Dev & Engineering ✓ Microsoft · Official

Azure Key Vault Certificates for Rust

Guides Rust applications in creating, managing, and using X.509 certificates in Azure Key Vault.

Dev & Engineering ✓ Microsoft · Official

Azure Conversation Understanding for Python

Helps Python developers analyze conversation intents and entities with the Azure SDK.

Design & Frontend ✓ Microsoft · Official

Frontend Design Review

Review or create production-ready frontend interfaces with stronger usability, craft, accessibility, and visual character.

Automation & Ops ✓ Microsoft · Official

Azure OpenAI Capacity Finder

Find available Azure OpenAI model capacity across regions and projects before choosing a deployment location.

Dev & Engineering ✓ Microsoft · Official

DebugView CLI Capture

Capture, filter, and analyze Windows application and kernel debug output from the command line.

Dev & Engineering ✓ Microsoft · Official

Azure AI Content Safety for Java

Helps Java applications detect harmful text and images and manage custom blocklists.

Dev & Engineering ✓ Microsoft · Official

Azure Content Understanding for Python

Extract structured multimodal content from documents, images, audio, and video with Python.

Dev & Engineering ✓ Microsoft · Official

Azure AI Anomaly Detector for Java

A practical Java skill for building univariate and multivariate time-series anomaly detection with Azure AI Anomaly Detector.

Data & Analysis ✓ Microsoft · Official

Azure Data Explorer Kusto Analytics

Query, aggregate, and analyze Azure Data Explorer logs, telemetry, and time-series data with KQL.

Automation & Ops ✓ Microsoft · Official

Azure Deployment Preparation

Prepares azd-based Azure projects with deployment configuration, infrastructure code, and container files.

Dev & Engineering ✓ Microsoft · Official

M365 Agents Toolkit Installer

Install or update the M365 Agents Toolkit CLI and VS Code extension with intent-aware commands.

Related skills