Skills for everyday work.

Nov 3

SQL Object Impact Analysis

Skill update · Coding

Nov 2

Security Review

Skill update · Coding

October 20267 reads

Oct 22

PHP Version Checker Skills

Skill update · Coding

Oct 20

java-upgrade

Skill update · Coding

Oct 15

batch-script-progress

Skill update · Coding

Oct 13

python-mcp-server-generator

Skill update · Coding

Oct 8

foundry-hosted-agent-copilotkit

Skill update · Coding

Oct 6

claude-api

Skill update · Coding

Oct 2

microsoft-store-publisher

Skill update · Coding

September 20267 reads

Sep 24

code-review

Skill update · Coding

Sep 23

api-breaking-change-detector

Source available by · Coding

Sep 15

agent-skills

Skill update · Coding

Sep 14

anti-ui-slop

Source available by · Design · Added Sep 23

Sep 8

Copilot Community Audit Skills Tighten Evidence Rules for Docs and Tests

Skill update · Coding

Sep 5

gh-fix-ci

Source available by · Coding · Added Sep 23

Sep 3

Combine Matt Pocock’s skills for planning, tests and review

Weekly guide · Coding

August 20264 reads

Aug 27

Design useful agent tools with mcp-builder

Weekly guide · Coding

Aug 20

Check a local web app with webapp-testing

Weekly guide · Coding

Aug 13

Keep formulas and templates intact with the xlsx skill

Weekly guide · Office

Aug 6

Bring shared team rules into coding sessions

Weekly guide · Coding

July 20265 reads

Jul 30

Make technical instructions clearer with SimpleEnglish

Weekly guide · Office

Jul 23

Give an interface a clear direction with frontend-design

Weekly guide · Design

Jul 16

Create and revise Word files with the docx skill

Weekly guide · Office

Jul 9

Use GeoSQL to check spatial queries on a map

Weekly guide · Coding

Jul 2

Prepare video frames and transcripts with claude-real-video

Weekly guide · Media

Dates identify publication or directory addition, except rows labeled “Source available by.” Those retrospective dates mark when the cited version already existed; the directory shows when TechReadly added and reviewed it.

Find a tool for the job

Skill directory

32 skills, organized by what you want to do.

OfficeDocuments, spreadsheets & presentations5 skills

docx

Create, revise, and inspect Word documents with professional formatting.

Compatibility & details

Best for: Reports, proposals, briefs, and document QA
Works with: Claude Code and Agent Skills-compatible hosts
Source owner: anthropics

xlsx

Build and analyze spreadsheets while preserving formulas and layout.

Compatibility & details

Best for: Models, trackers, data cleanup, and spreadsheet QA
Works with: Claude Code and Agent Skills-compatible hosts
Source owner: anthropics

pptx

Create and edit slide decks with reusable layouts and visual checks.

Compatibility & details

Best for: Presentations, pitch decks, and internal reviews
Works with: Claude Code and Agent Skills-compatible hosts
Source owner: anthropics

pdf

Read, generate, transform, and visually verify PDF documents.

Compatibility & details

Best for: Research packets, forms, reports, and exports
Works with: Claude Code and Agent Skills-compatible hosts
Source owner: anthropics

SimpleEnglish

Apply controlled-language guidance to technical instructions.

Compatibility & details

Best for: Documentation, release notes and error messages
Works with: Agent Skills-compatible hosts
Source owner: AminBlg
Source snapshot: September 8, 2026
Added to TechReadly: September 8, 2026

CodingDevelopment, testing & agent tools20 skills

Docs Sync Audit

Compatibility & details

Best for: Maintainers checking setup guides, API references, CLI instructions and release documentation. This is a source-based introduction; no repository audit was performed.
Works with: Skill-capable coding agents with repository read access and search tools. Its bundled Python checker uses the standard library and flags command, link and environment-name mismatches; findings still need review. MIT-licensed instructions and code, with model costs separate. Broad audits must disclose areas not deeply inspected.
Source owner: github
Source snapshot: November 5, 2026
Added to TechReadly: November 5, 2026

SQL Object Impact Analysis

Compatibility & details

Best for: Developers changing tables, columns, procedures, views or triggers in unfamiliar legacy code. This is a source-based introduction; no database trace was performed.
Works with: Skill-capable coding agents with access to the relevant SQL scripts and application files. MIT-licensed instructions; model and tool costs are separate. Static searches can miss dynamic SQL and reflection, so the report must state its limits and does not guarantee a safe migration.
Source owner: github
Source snapshot: November 3, 2026
Added to TechReadly: November 3, 2026

Security Review

Compatibility & details

Best for: Developers reviewing an existing codebase who want findings tied to files, code snippets, severity and confidence. This is a source-based introduction; no security scan was performed.
Works with: Coding agents that can load the repository’s skill instructions and inspect source files. The MIT-licensed instructions include language and vulnerability references; model usage is separate. Treat the package watchlist as a starting point, and verify findings against current advisories and actual exploit paths.
Source owner: github
Source snapshot: November 2, 2026
Added to TechReadly: November 2, 2026

PHP Version Checker Skills

Two Skills help an AI agent inventory PHP patch status without exploit probes. phpcheck-url reads advertised versions from HTTP headers, the labels sent with a web response; phpcheck-server directs the agent to verify the actual PHP-FPM or Apache build and its package advisory. The bundled checker can return OK, VULNERABLE or UNKNOWN and emit JSON.

Compatibility & details

Best for: Website maintainers triaging the October 21 PHP security release. A version comparison is a starting point: hidden headers, vendor backports and differences between command-line PHP and the serving runtime require further checks. UNKNOWN does not mean safe. The Skills do not install updates or independently prove remediation.
Works with: A host that supports SKILL.md instructions. The URL checker uses Node.js with built-in fetch and core modules, without npm dependencies; server checks require access to the relevant runtime and package information. Apache-2.0 source. The new repository received fixes on October 21; external code contributions are not accepted, though bug reports and non-code feedback are welcome.
Source owner: openai
Upstream update: October 21, 2026
Source snapshot: October 22, 2026
Added to TechReadly: October 22, 2026

java-upgrade

Guides an AI coding agent through Java 11→17, 17→21 or 21→25 migrations. It reads Maven, Gradle and CI configuration, selects the relevant version guide, updates the build before adopting language features, and calls for builds and tests at each step. The October 15 change replaces large always-on instruction files with an on-demand Skill and version-specific references.

Compatibility & details

Best for: Java maintainers who want small, reviewable upgrade steps and explicit framework-compatibility checks. The guide covers behavior preservation and keeps preview features opt-in. It provides migration instructions; it is not an independently tested guarantee that a particular application will upgrade successfully.
Works with: A host that supports GitHub Copilot Agent Skills, the target JDK and the project’s Maven or Gradle build. OpenRewrite can assist when available. The separate Microsoft modernize-java plugin is optional for a more automated workflow. Repository content is MIT-licensed; model-host fees and project dependencies are separate.
Source owner: github
Upstream update: October 15, 2026
Source snapshot: October 20, 2026
Added to TechReadly: October 20, 2026

batch-script-progress

The `batch-script-progress` Skill provides a specification for adding startup notices, bounded progress updates, sanitized per-item failure diagnostics, and terminal summaries to automation scripts that iterate through records, files, or API endpoints. Its stated design goal is to deliver runtime visibility without altering the existing processing semantics, exit statuses, retry boundaries, or stdout data contracts of the underlying code. The specification preserves stdout for its established result or protocol output, defaulting diagnostic messages to stderr unless callers reject that channel, in which case developers must use an existing logger or halt if no safe stream is available. In its accounting model, the Skill enforces that unknown totals remain explicitly unknown when streams are lazily evaluated, strictly barring the invention of artificial ETAs, completion percentages, or premature stream materialization. Terminal counts follow an accounting rule where processed items equal the sum of succeeded, failed, and skipped items, counting each unit once regardless of internal retry attempts. In-flight tasks cannot be marked successful before durable commits occur, rolled-back transactions must be logged as rolled back rather than completed, and scripts that traditionally return exit code zero during partial failures must preserve that exact exit status while emitting an honest terminal summary.

Compatibility & details

Best for: This Skill is targeted at software developers, automation engineers tasked with maintaining or generating long-running command-line scripts across multi-language production environments. It is specifically framed for teams refactoring silent loops, bulk data processors, and batch utilities that need diagnostic observability while integrating strictly into downstream pipelines, continuous integration runners, or orchestration managers that depend on rigid exit codes. The guidance explicitly excludes interactive UI progress bars, conversational AI agent progress messages, and distributed infrastructure monitoring tools. It is tailored for practitioners who need consistent logging vocabulary across disparate runtimes while respecting runtime-specific constraints, such as warning Node.js users that synchronous blocking prevents timer-based heartbeats, and reminding PowerShell users that `Write-Output` pollutes pipeline streams.
Works with: Operating the Skill inside an AI assistant requires an environment that supports the GitHub Copilot Agent Skills format. The repository operates under an MIT license, and the Skill itself does not mandate any external third-party service dependencies or proprietary API keys, though underlying provider host model fees remain separate. Running the repository's cross-language probe verifier (`tests/verify-examples.mjs`) mandates three local runtimes: Node.js 20 or later, Python 3.9 or later, and PowerShell 5.1 or later. The bundled verification script fails explicitly if any runtime is missing rather than skipping test coverage. The verifier uses harmless fixtures to confirm liveness by reading streaming output while child processes remain active, rather than evaluating buffered logs post-exit, though the repository documentation notes that bundled probe passes do not guarantee arbitrary production script validity.
Source owner: github
Upstream update: October 14, 2026
Source snapshot: October 15, 2026
Added to TechReadly: October 15, 2026

python-mcp-server-generator

The python-mcp-server-generator skill instructs GitHub Copilot Agent Skills hosts to scaffold, implement, and configure a Python-based Model Context Protocol (MCP) server that exposes local Python functions, existing APIs, or databases as structured tools for AI clients. The skill generates an application project layout using uv with --no-package, installs the mcp[cli]>=2,<3 package alongside pytest, and writes a standardized server.py using typed schemas and docstrings to automatically generate tool input definitions. It configures stdio as the primary local transport, supports optional Streamable HTTP references, establishes in-memory Client integration tests via anyio without requiring subprocesses or network ports, and outputs client connection manifests for Visual Studio Code (.vscode/mcp.json) and Claude Desktop (uv run mcp install server.py).

Compatibility & details

Best for: This skill is designed specifically for Python developers using a GitHub Copilot Agent Skills host to expose local code, data stores, or REST APIs to MCP-compliant AI hosts such as Claude Desktop or VS Code, without building custom protocol serialization from scratch. It targets new servers using MCP specification revision 2026-07-28 and the SDK v2 MCPServer class. The skill explicitly excludes creating MCP client applications and does not support scaffolding servers written in languages other than Python.
Works with: The skill requires Python 3.10 or later, the uv packaging manager, network access to resolve dependencies, and a host environment supporting GitHub Copilot Agent Skills; running the optional MCP Inspector debugging tool (uv run mcp dev server.py) requires Node.js. If exposed tools query external services or databases, those services may need their own credentials and access configuration, though the skill itself imposes no blanket paid API tier. The skill is published under the MIT license within the github/awesome-copilot repository (commit 3fb8ae0, October 9, 2026) and uses MCP Python SDK v2. SDK v1 examples are incompatible with this template: v2 uses MCPServer instead of FastMCP, and remote servers use Streamable HTTP rather than SSE.
Source owner: github
Upstream update: October 9, 2026
Source snapshot: October 13, 2026
Added to TechReadly: October 13, 2026

foundry-hosted-agent-copilotkit

Connects a CopilotKit React or Next.js frontend to a Microsoft Agent Framework agent deployed as a Microsoft Foundry hosted agent through an intermediary AG-UI server gateway. The revised workflow uses native Foundry conversations, durable tool approvals and background execution. Hosted agents do not speak AG-UI directly; the default gateway connects to their Responses endpoint and keeps Microsoft Entra ID authentication on the server.

Compatibility & details

Best for: Frontend and full-stack developers building React or Next.js user interfaces that require conversational copilot components alongside hosted Microsoft Agent Framework backends. It covers human-in-the-loop tool approvals, generative UI, deployment and event-stream debugging. Shared-state support depends on the selected wiring and must be configured and tested.
Works with: Requires an active Azure subscription with a Microsoft Foundry project configured for the paid hosted agent preview, supported by Node.js and Python runtimes. Requires GitHub Copilot with Agent Skills support, Azure CLI and Azure Developer CLI (azd) with the azure.ai.agents extension. The plugin bundles Microsoft Learn MCP and Foundry MCP; Foundry MCP requires Microsoft Entra ID sign-in. MIT license for the repository; Azure compute and model usage are billed separately.
Source owner: github
Upstream update: October 7, 2026
Source snapshot: October 8, 2026
Added to TechReadly: October 8, 2026

claude-api

Configuring, provisioning, and validating Claude Platform Managed Agents from bundled console templates or remote documentation URLs.

Compatibility & details

Best for: Developers working with Claude Code or the Claude API who need to bootstrap Managed Agent configurations, environment boundaries, and credential vaults without authoring setup schemas from scratch.
Works with: Anthropic ant CLI version 1.34.0 or higher, Claude Code, and Anthropic API workspaces with active billing. Claude Platform on AWS uses an SDK fallback because the CLI has no SigV4 mode; this does not establish support on Bedrock, Vertex AI, or Foundry. Apache 2.0 license for this skill directory.
Source owner: anthropics
Upstream update: October 5, 2026
Source snapshot: October 6, 2026
Added to TechReadly: October 6, 2026

microsoft-store-publisher

Prepare Windows desktop MSIX submissions for Microsoft Partner Center: capture product identity, inspect packages, run the Windows App Certification Kit, and assemble listing and certification notes. Binding agreements and certification submissions require action-specific approval; the skill does not guarantee certification.

Compatibility & details

Best for: Windows desktop developers preparing a first MSIX release through Partner Center.
Works with: GitHub Copilot with Agent Skills support; Windows packaging tools, Windows App Certification Kit, and a Partner Center developer account. The optional Edge automation helper uses Node.js and Playwright with interactive sign-in. MIT license.
Source owner: github
Upstream update: September 30, 2026
Source snapshot: October 2, 2026
Added to TechReadly: October 2, 2026

code-review

Guide reviewers through submission standards for contributions to github/awesome-copilot, including fit, value, provenance and AI disclosure, differentiation, supporting evidence, marketing claims, and maintainability alongside deterministic checks.

Compatibility & details

Best for: Maintainers and contributors evaluating submissions to the github/awesome-copilot repository.
Works with: GitHub Copilot in github/awesome-copilot; wired into the repository Copilot instructions and paired with its deterministic checks.
Source owner: github
Upstream update: September 23, 2026
Source snapshot: September 24, 2026
Added to TechReadly: September 24, 2026

api-breaking-change-detector

Compare C# API contracts with JavaScript or TypeScript consumers to find likely breaking changes.

Compatibility & details

Best for: Reviewing controller or DTO changes before merging a full-stack pull request
Works with: GitHub Copilot; C# Web API and JavaScript or TypeScript clients
Source owner: github
Upstream update: September 22, 2026
Source snapshot: September 23, 2026
Added to TechReadly: September 23, 2026

gh-fix-ci

Inspect failing GitHub Actions pull-request checks and their logs before planning a fix.

Compatibility & details

Best for: Triaging failed GitHub Actions checks on a pull request
Works with: Codex and an authenticated GitHub CLI; external CI providers are outside scope
Source owner: openai
Upstream update: January 30, 2026
Source snapshot: September 23, 2026
Added to TechReadly: September 23, 2026

test-gap-audit

Review existing assertions and identify missing regression test scenarios.

Compatibility & details

Best for: Prioritizing tests for a feature or pull request
Works with: GitHub Copilot with Agent Skills support
Source owner: github
Upstream update: September 7, 2026
Source snapshot: September 8, 2026

webapp-testing

Inspect and test local web applications through browser automation.

Compatibility & details

Best for: UI verification, regression checks, and debugging
Works with: Claude Code and Agent Skills-compatible hosts
Source owner: anthropics

mcp-builder

Design MCP servers that expose reliable tools to AI agents.

Compatibility & details

Best for: Agent integrations, internal tools, and API wrappers
Works with: Claude Code and MCP-compatible clients
Source owner: anthropics

agent-skills

Give coding agents structured workflows for specification, planning, implementation, testing, review, and shipping.

Compatibility & details

Best for: Teams standardizing software lifecycle practices across coding-agent hosts
Works with: Claude Code, Cursor, Gemini CLI, OpenCode, GitHub Copilot, Codex, and other hosts documented by the repository
Source owner: addyosmani
Upstream update: September 12, 2026
Source snapshot: September 15, 2026

GeoSQL

Generate spatial SQL and inspect results on a map.

Compatibility & details

Best for: Geospatial analysis across supported databases
Works with: Claude Code, Codex, GitHub Copilot
Source owner: dekart-xyz
Source snapshot: September 8, 2026
Added to TechReadly: September 8, 2026

adlc-team-skills

Load shared team rules and decisions when coding tasks need them.

Compatibility & details

Best for: Teams maintaining agent conventions in Git
Works with: Claude Code, Codex, OpenCode and compatible hosts
Source owner: tikalk
Source snapshot: September 8, 2026
Added to TechReadly: September 8, 2026

matt-pocock-skills

Combine requirements interviews, test-driven work and code review.

Compatibility & details

Best for: Developers shaping repeatable engineering workflows
Works with: Claude Code, Codex and compatible agents
Source owner: mattpocock
Source snapshot: September 8, 2026
Added to TechReadly: September 8, 2026

DesignInterfaces, diagrams & visual assets5 skills

anti-ui-slop

Ground web or iOS interface work in the product brief, existing design system, and one focused playbook.

Compatibility & details

Best for: Designing, refining, or reviewing an interface with an AI coding agent
Works with: Claude Code, Codex, Cursor, and GitHub Copilot; optional paid UIZZE MCP
Source owner: github
Upstream update: August 24, 2026
Source snapshot: September 23, 2026
Added to TechReadly: September 23, 2026