# WorkBuddy Skill Atlas

> Search public Agent Skills and adapt reviewed workflows for WorkBuddy. Every catalog record links to an immutable GitHub commit. Indexed content is not automatically trusted or installed.

The public catalog size is authoritative in `catalog-meta.json` under `records`;
use it to narrow a search, then inspect the linked source, license, and permissions before installation.

Release ZIPs are accompanied by a `SHA256SUMS` asset for offline integrity verification.

For a local checkout, use `python3 scripts/query_catalog.py <term> --installable --limit 10`
to shortlist reviewed packages with no known static flags, primary-looking sources,
score >= 80, and unique results. This is a shortlist only; inspect provenance,
license, permissions, and side effects before importing.

## Use the catalog

- Web search: https://sandbaseai.github.io/workbuddy-skill/
- Chinese search: https://sandbaseai.github.io/workbuddy-skill/zh-CN.html
- Compact JSON catalog: https://sandbaseai.github.io/workbuddy-skill/catalog.json
- Compact catalog JSON Schema: https://sandbaseai.github.io/workbuddy-skill/catalog-schema.json
- Reviewed package JSON: https://sandbaseai.github.io/workbuddy-skill/packages.json
- Human-readable reviewed package index: https://sandbaseai.github.io/workbuddy-skill/packages.html
- Shareable category directory: https://sandbaseai.github.io/workbuddy-skill/categories.html
- Chinese category directory: https://sandbaseai.github.io/workbuddy-skill/categories.zh-CN.html
- Reviewed package JSON Schema: https://sandbaseai.github.io/workbuddy-skill/packages-schema.json
- Each reviewed package record includes `source_url`, pinned `sha`, `download_url`, `asset`, `download_command`, and `checksum_url`.
- If a query has no catalog match, use the Atlas empty-state link or [GitHub's current `SKILL.md` code search](https://github.com/search?q=filename%3ASKILL.md&type=code); this is discovery only and does not install or trust external code automatically.
- Catalog metadata: https://sandbaseai.github.io/workbuddy-skill/catalog-meta.json
- Catalog metadata JSON Schema: https://sandbaseai.github.io/workbuddy-skill/catalog-meta-schema.json
- Catalog metadata includes `catalog_sha256`, a SHA-256 fingerprint of the published catalog bytes.
- Source repository: https://github.com/sandbaseai/workbuddy-skill
- Citation metadata: https://github.com/sandbaseai/workbuddy-skill/blob/main/CITATION.cff
- Curated WorkBuddy ecosystem index: https://github.com/sandbaseai/awesome-workbuddy
- MIT Superpowers agentic software-development methodology (external process reference): https://github.com/obra/superpowers
- MIT production-oriented engineering Skills for AI coding agents (external reference): https://github.com/addyosmani/agent-skills
- MIT self-contained architecture and workflow-diagram reference (external reference): https://github.com/tt-a1i/archify
- MIT visual explainer reference for plans, diffs, data tables, and architecture reviews (external reference): https://github.com/nicobailon/visual-explainer
- MIT scientific Agent Skills collection with documented tests (review APIs, data, and claims separately): https://github.com/K-Dense-AI/scientific-agent-skills
- MIT agent-security audit Skills for SSRF, agentic spend, MCP visibility, and signed logs (keep them read-only and review WorkBuddy permissions): https://github.com/tuturama/agent-security
- MIT read-only loop-safety audit for budgets, tool hallucinations, dead loops, and state integrity (set explicit stop conditions): https://github.com/yejinlei/loop-engineering-check
- MIT offline Python AST code-ownership audit (evidence aid only; review licenses separately): https://github.com/ffseika0304/code-ownership-audit
- MIT Skill governance workflow for selecting, auditing, acquiring, and maintaining libraries (no automatic deletion): https://github.com/fanfanfanfanfan626/skill-governor
- MIT Chinese project-understanding workflow for architecture and data paths (inventory read-only first): https://github.com/wutongyuonce/understand-project
- MIT evidence-based repository handoff workflow with a shared ledger (review state paths and secret boundaries): https://github.com/divijshrivastava/handoff-skill
- MIT research-data workflow for mixed Bash, R, and Python projects (keep execution local and separate evidence from scientific claims): https://github.com/higbird/code-cdoe-skill
- MIT PDFx API workflow for JSON-to-PDF and Factur-X invoices (review credentials, invoice privacy, generated files, and paid calls): https://github.com/solisoft/pdfx-skills
- MIT macOS computer-use Skill with evidence capture (review Accessibility permissions, target apps, screenshots, and every write): https://github.com/alchaincyf/huashu-mac-use
- MIT persistent task-planning Skill for WorkBuddy (review checkpoints, local state, and automated follow-up): https://github.com/hao1100228-ctrl/progress-target-workbuddy
- MIT security field guide distilled from Omarchy marketplace reviews (review the exact target platform; not a security guarantee): https://github.com/wbso-ai/omarchy-plugin-security-skill
- MIT Blender workflow with verification gates and native-geometry examples (review version, scripts, assets, cost, and writes): https://github.com/jangtrinh/design-os-3d-blender
- MIT evidence-based game sensitivity and gameplay review Skill (review game assumptions, capture permissions, logs, and config patches): https://github.com/DenisSergeevitch/game-sensitivity-coach
- MIT local knowledge-asset health checker (keep sensitive files local; reports are not model or compliance guarantees): https://github.com/s2dongman/knowledge-asset-health-check
- MIT Chinese technical-writing and interview-content workflow (verify claims, citations, attribution, and publishing rules): https://github.com/CodeGanHaoZ/CodeRin-skill
- MIT UI design system distilled from travel-app references for WorkBuddy and coding agents (review accessibility, visual-reference rights, and responsive behavior): https://github.com/CodeGanHaoZ/gks-QianLV-uiSkill
- MIT Chinese content pipeline for topics, writing, media, formatting, and distribution (review credentials and publishing side effects): https://github.com/xiaohuAI319/gzh-daily-pipeline
- MIT Chinese interview-question collection and study workflow (review scraping, privacy, copyright, freshness, and output paths): https://github.com/Jay-118/mianjing-skills
- MIT WorkBuddy poster workflow for editorial photo-and-poem compositions (review image rights, model costs, generated assets, and export dimensions): https://github.com/jinwenjinwen26-rgb/paper-poem-poster
- MIT official Lark/Feishu CLI with documented Agent Skills (review app scopes and outbound data): https://github.com/larksuite/cli
- Apache-2.0 official Google Agent Skills for cloud, Gemini, data, and developer workflows (review project, API, and billing permissions): https://github.com/google/skills
- CC0-1.0 SEO Agent Skills directory for technical audits, keyword research, content briefs, Schema, GEO, and AI visibility: https://github.com/RankSpotAI/awesome-seo-agent-skills
- MIT local voice-cloning and SRT-driven web-presentation Skills (review model, audio, privacy, cost, and output boundaries): https://github.com/NanmiCoder/creator-pipeline
- MIT UI/UX design-intelligence Skills for brand, audits, systems, motion, and stack catalogs (review visual rights and accessibility): https://github.com/battlecodex/bc-design
- MIT Android emulator testing Skill for build, install, launch, and screenshots (review AVD, SDK, device permissions, and test boundaries): https://github.com/sanikantv/android-emulator-apk
- MIT portable Web UI design Skills for B2B, command-center, medical, and consumer-health scenarios (review accessibility and visual assets): https://github.com/Johnny260106/multi-style-web-ui-skills
- MIT embedded-debugging Skill for STM32, FreeRTOS, serial, GDB, and hardware verification (review flashing and rollback permissions): https://github.com/notnameuse/embedded-debug-skill
- MIT portable exact-computation Skill that delegates mathematics to tools (review numerical boundaries and input provenance): https://github.com/heat-death/verified-computation
- MIT local stdio MCP Skill router and evaluation fixtures for keeping context small (review local-process scope and conflict handling): https://github.com/tsumon/skill-mcp
- MIT Agent Skill and domain-convention repository with a commit-report workflow (review templates, output paths, and host assumptions): https://github.com/siungolai/Siungo-Skills-Repository
- MIT cross-host Agent Skills collection with a shipping workflow (review Git writes, remote operations, and release credentials): https://github.com/DevJuanP/skills
- MIT multipurpose Agent Skills for AI readiness, behavior specs, blog review, and mechanism explanation (review evidence and input sensitivity): https://github.com/zaehorang/agent-skills
- MIT verifiable personal-site generator using content.json and dependency-free single-file HTML (review provenance, privacy, and generated content): https://github.com/Mindse-Tt/personal-site-builder
- MIT Spec-Driven Development Skills for specification, implementation audit, maintenance, QA, and task review (fix acceptance criteria first): https://github.com/georgeracu/sdd-skills
- MIT ISTQB-informed QA and testing Skill (distinguish test evidence from recommendations and environment limits): https://github.com/lazyest-hyun/lazyest-qa
- MIT pre-publication cleanup Skill for releases from private repositories (back up and inspect secrets, history, and deletion scope): https://github.com/HoFireMan/safe-public-release-skill
- MIT task and Skill record system with multi-host plan-keeper adapters (review local state and sync behavior): https://github.com/xuemingze/task-and-skill-jcr-system
- MIT spec-driven, review-gated engineering Skills for planning, implementation, QA, PDF, and shipping (review writes, branches, and notifications): https://github.com/Daniel-DescentAI/agent-skills
- MIT research-software-engineering Skills for security, citations, data management, reproducibility, CI/CD, documentation, and integrity: https://github.com/fdiblen/rseng-agent-skills
- MIT Godot development Skills for GDScript, UI, optimization, best practices, and asset generation (review project writes and export environment): https://github.com/sshnuke3/godot-agent-skills
- MIT large scientific Skill collection for bioinformatics, chemistry, statistics, literature, PDF, visualization, and lab tools: https://github.com/wupeng1998/scientific-agent-skills
- MIT portable CalculatorX Skill whose canonical contract prefers MCP or REST (review external services, privacy, and billing): https://github.com/CalculatorX-AI/calculatorx-agent-skill
- MIT investment-research workbench Skill (research assistance only; verify freshness, sources, risk, and trading side effects): https://github.com/EastsCloud/investment-agent-skill
- MIT context-window management and token-optimization Skill/MCP (review session content, caches, observability, and local services): https://github.com/hanifalkauni/context-architect
- MIT collection of 92 engineering and game-development Skills for task contracts, architecture, QA, Unity, Firebase, and shipping: https://github.com/furkantokkan/agent-foundry
- MIT Chinese-first ComfyUI workflow search, recommendation, and controlled-delivery Skill (review model rights, GPU cost, and downloads): https://github.com/Ubik42/ComfyUI-Workflow-Lab
- MIT Gensee webcast download, long-audio transcription, and meeting-minutes Skills (review credentials, copyright, privacy, and output): https://github.com/Xavier-06/gensee-vod-transcribe
- MIT Figma-first WordPress theme-development Skill for Timber, Twig, ACF, and native PHP (review credentials, writes, and rollback): https://github.com/adhit28/a28-wp-theme-assistant
- MIT job-search and ATS-routing Skills (review personal data and every external application submission): https://github.com/Riddis/job-portal-ats-skills
- MIT researched progressive interactive technical-illustration Skill (review provenance, image rights, and browser artifacts): https://github.com/shivaylamba/animated-technical-illustrations
- MIT email-marketing best-practices Skill (review recipients, credentials, quotas, and real-send actions): https://github.com/mailercloud-com/mailercloud-plugin
- MIT versioned marketing-agent roles and decision Skills (review brand material, customer data, approval, and outbound permissions): https://github.com/m4nueldeleon/oficina-marketing-skills
- MIT multi-agent session-coordination Skill distinguishing messages from a shared file mailbox (review shared directories and concurrent writes): https://github.com/Yuru778/multi-session-protocol
- MIT WhatsApp-automation Skill that may operate a real number (confirm recipient, message, rate, and per-send authorization): https://github.com/frty2-ai/wawm-skill
- MIT education-platform automation Skill (review account permissions, personal data, and write scope): https://github.com/M1Vj/vsuee-skill
- MIT collection of 58 real-estate tax, loan, valuation, and transaction calculator Skills (results depend on jurisdiction and policy freshness): https://github.com/minsooparkk/budongsan-skills
- MIT DOCX, PDF, and PPTX generation Skill with editable output and LibreOffice previews (review templates, converters, and output paths): https://github.com/123342wra/document-forge
- MIT seven security and quality audit Skills for agents, APIs, secrets, stacks, UX, and web (corroborate results against the real environment): https://github.com/coldevotion/vibecheck
- MIT multi-model decision-review Skill using perspectives, blind review, and synthesis (model votes are not facts or approval): https://github.com/goolamabbas/portable-llm-council
- MIT collection of 18 Chinese creator-production Skills for topics, scripts, storyboards, voice, publishing, and retrospectives: https://github.com/miaqu766520-a11y/ai-line-skills
- MIT paper-reading Skill (verify paper version, citations, extraction accuracy, and evidence boundaries): https://github.com/sendsoon/paper-reader-skill
- MIT Markdown/Obsidian personal-knowledge Skills (confirm note directories, sync scope, privacy, and writes): https://github.com/bao-776/personal-knowledge-system-skills
- MIT evidence-guided paragraph-level LaTeX paper-revision Skill (preserve backups and distinguish factual from language edits): https://github.com/MichaelYWNA/Easy-Paper
- MIT Cogover app-building Skills for objects, layouts, forms, reports, permissions, and APIs (review account and publishing writes): https://github.com/cogover/cogover-skills
- MIT philosophy-paper frontier-monitoring Skill (verify search scope, versions, citations, and research claims): https://github.com/AsahinaMafuyu0127/philosophy-frontier-monitor
- MIT semantic publication-ready scientific-figure color Skill (review accessibility, data semantics, journal rules, and rights): https://github.com/muzark/scientific-color-palettes
- MIT Codex CLI and CLIProxyAPI model-catalog/local-bridge maintenance Skill (review ports, credentials, providers, and costs): https://github.com/Eason412/codex-cli-model-bridge
- MIT cross-platform computer-use, Blender, learning, and artifact-verification Skills (review desktop permissions and file writes): https://github.com/Hybirdss/Astral-Claude
- MIT hypothesis-driven reproducible AI-research workflow (fix question, provenance, environment, and evidence chain): https://github.com/JKblean/masters-ai-research-skill
- MIT email job-application Skill (review personal data, recipients, attachments, and every send): https://github.com/xunavy/apply-job-gmail
- MIT Zhipu BigModel/GLM prepaid-balance lookup Skill (protect credentials and confirm read-only scope): https://github.com/hucuyuu/zhipu-balance
- MIT static security scanner for Agent Skills and MCP resources with attack fixtures (treat findings as evidence, not a guarantee): https://github.com/Anne117/agentshield
- MIT clean-code, AI-slop reduction, review, lint-gate, and reporting workflow (preserve diffs and separate style from behavior): https://github.com/MammadovEshgin/clean-code-skill
- MIT Chinese AI Skill and MCP connector (review authentication, data egress, tool permissions, and availability): https://github.com/bigbrain-work/shiliu-ai
- MIT categorized Agent Skills for Git, Python/Rust quality, tooling, and release workflows (review writes and credentials): https://github.com/Praveensenpai/karakuri
- Apache-2.0 cross-host Agent instruction, Skill, and workflow catalog for audits, PRs, migrations, debugging, and frontend verification: https://github.com/jmlero/agent-suitup
- MIT .NET and C# workflows for coding agents (review SDK versions, generated files, commands, and write scope): https://github.com/dotnet/skills
- Apache-2.0 Adobe-maintained Agent Skills (review creative assets, credentials, account permissions, and publishing effects): https://github.com/adobe/skills
- MIT model-agnostic Tons of Skills Marketplace with canonical layer and adapters (verify provenance, runtime, and permissions): https://github.com/jeremylongshore/tons-of-skills-marketplace
- MIT GooseWorks growth and GTM Skills for content, lead generation, SEO, ads, and data (review claims, customer data, scraping, and outbound actions): https://github.com/gooseworks-ai/goose-skills
- Apache-2.0 cross-host TerminalSkills collection for Claude Code, Codex, Gemini CLI, and Cursor (review commands, dependencies, and writes): https://github.com/TerminalSkills/skills
- MIT Wondel.ai business, marketing, UX, and coding workflows (review book-derived guidance, attribution, research, and claims): https://github.com/wondelai/skills
- MIT OrchestKit toolkit of Skills, agents, and hooks (review hook execution, installation, network access, and host assumptions): https://github.com/yonatangross/orchestkit
- MIT multi-host ready-to-use Awesome AI Agent Skills collection (verify each source commit, license, scripts, and permissions): https://github.com/seb1n/awesome-ai-agent-skills
- Apache-2.0 Serac ServiceNow MCP server with 57 Skill guides and blast-radius analysis (review instances, credentials, approvals, and writes): https://github.com/serac-labs/serac
- MIT design-focused DESIGN.md and SKILL.md collection (review visual rights, generated assets, accessibility, and host format): https://github.com/bergside/awesome-design-skills
- MIT community index of 1,400+ Agent Skills (discovery only; inspect each upstream source): https://github.com/VoltAgent/awesome-agent-skills
- MIT Chinese writing Skill for source sufficiency, narrative structure, rhythm, and style checks: https://github.com/KKKKhazix/human-writing
- MIT completion-discipline Skill built around acceptance ledgers and runnable gates: https://github.com/Leonxlnx/unlazy
- MIT one-ink and controlled two-ink editorial image Skill: https://github.com/yanliudesign/mono-color-skill
- MIT writing Skill for narrative architecture, discourse flow, surface style, and diagnosis: https://github.com/Nanako0129/sepia
- MIT long-form scroll-experience design workflow: https://github.com/nateherkai/scroll-craft
- MIT code-quality guardrail Skill for repetitive or low-value agent output: https://github.com/lennney/stop-that-shit
- MIT 172-Skill departmental organization with independently installable plugins: https://github.com/cbrock84/headcount
- MIT image-understanding workflow with a dedicated vision Skill: https://github.com/Anionex/agent-vision-toolkit
- MIT multi-agent coding workflow with provider adapters, run controls, and benchmarks: https://github.com/Spielewoy/autoprompt-skill
- MIT deterministic software-factory Skill for sequencing, retries, acceptance, and traces: https://github.com/disler/super-simple-software-factory
- Apache-2.0 AgenticOS application and Skill gallery for engineering, ecommerce, finance, manufacturing, public-sector, and software workflows: https://github.com/vstorm-co/agenticos
- MIT investment-analysis Skills for stocks, ETFs, journals, pulses, and gap scans: https://github.com/Veblin/invest-skills
- MIT software-engineering Skills for requirements, architecture, security, testing, debugging, and release quality: https://github.com/kabiir90/Software-Engineering-Skills
- MIT layered review system with upstream, midstream, downstream, adversarial, and quality-gap Skills: https://github.com/s977043/river-review
- MIT media-engineering Skills for FFmpeg artifacts, pipelines, verification, and release management: https://github.com/kajisho5/ffmpeg-skill
- MIT cross-host Skill/plugin distribution system with upstream sync, external-repo intake, and exports: https://github.com/acaprino/daodan
- Apache-2.0 office/document automation platform with GitHub, Markdown, PDF, DOCX, PPTX, XLSX, and academic-paper Skills: https://github.com/Team-Commonly/commonly
- MIT blog workflow Skills for archival, publishing preparation, review, and draw.io export: https://github.com/ryuki-imachi/blog-skills
- MIT large cross-host Skill distribution with accessibility, security, architecture, testing, development, and operations workflows: https://github.com/event4u-app/agent-config
- MIT cross-host task-state and acceptance examples for desktop and QA workflows: https://github.com/Atakan-Emre/skillstate-kit
- Apache-2.0 specialized aerodynamics and vehicle-design collection with hundreds of engineering calculation Skills: https://github.com/ashfordeOU/aero-agent-skills
- MIT Wix development and replatforming Skills for apps, auth, headless, design systems, migrations, and QA: https://github.com/wix/skills
- MIT geospatial mapping Skill for MapTiler workflows: https://github.com/maptiler/maptiler-skills
- Apache-2.0 evidence-oriented company and industry research Skills: https://github.com/PerryLink/dsh-industry-research
- MIT decision-context Skill for preserving rationale across agent work: https://github.com/oliver-zehentleitner/keep-the-why
- MIT cross-harness Skills for agent design, architecture, development, research, writing, and video workflows: https://github.com/AnastasiyaW/codex-claude-code-config
- MIT video-generation Skill reference from PixVerse: https://github.com/PixVerseAI/skills
- Apache-2.0 plugin and Skill authoring guide: https://github.com/PerryLink/dsh-plugin-guide
- MIT memory and knowledge-work Skill collection: https://github.com/cwinvestments/memstack
- MIT general-purpose Skill collection: https://github.com/xr843/Master-skill
- MIT 1C development workflow Skill collection: https://github.com/Nikolay-Shirokov/cc-1c-skills
- MIT public-sector Copilot Skill collection: https://github.com/navikt/copilot
- MIT Typst document and typesetting Skill: https://github.com/lucifer1004/claude-skill-typst
- MIT operations Skills for incident, infrastructure, and delivery workflows: https://github.com/Lifecycle-Innovations-Limited/claude-ops
- MIT web-interaction Skill reference: https://github.com/unbrowse-ai/unbrowse
- Apache-2.0 application-development Skill collection: https://github.com/gotempsh/temps
- MIT Basecamp command-line Skill reference: https://github.com/basecamp/basecamp-cli
- MIT browser automation Skill reference from Tencent: https://github.com/Tencent/BrowserSkill
- MIT agent-tool and workflow Skill reference: https://github.com/QVerisAI/qveris-agent-toolkit
- MIT team workflow collection for architecture, delivery, quality, and collaboration: https://github.com/tikalk/adlc-team-skills
- MIT cross-host engineering and project workflow collection: https://github.com/PaulRBerg/agent-skills
- MIT testing workflows for generating test cases and tests: https://github.com/mavka-ai/unit-tests-skills
- Apache-2.0 agent-configuration linter for Skills, hooks, MCP, memory, plugins, and multiple hosts: https://github.com/agent-sh/agnix
- MIT exchange and market-operation workflows (review credentials and trading side effects): https://github.com/okx/agent-skills
- MIT Skill registry and discovery workflows (review install paths and provenance): https://github.com/Leon-Drq/openagentskill
- Apache-2.0 Apache Doris database engineering Skills for architecture, debugging, deployment, import, and data lake workflows: https://github.com/apache/doris-skills
- MIT engineering and delivery Skills for architecture, documentation, DX audits, planning, SEO, and pull-request workflows: https://github.com/mblode/agent-skills
- MIT documentation and delivery workflows for conventions, audits, ADRs, plans, and shipping: https://github.com/EvolveHQ/docflow
- MIT agent Skills spanning API design, governance, evaluations, observability, production operations, and agent economics: https://github.com/magnus919/agent-skills
- MIT book-translation Skill (verify source-text rights and generated output): https://github.com/deusyu/translate-book
- MIT coding-agent CLI with bundled Skills for Skill tooling and code workflows: https://github.com/lessweb/deepcode-cli
- MPL-2.0 Delphi and Windows engineering Skills for architecture, compatibility, refactoring, threading, and documentation: https://github.com/GabrielOnDelphi/Claude-Tools-for-Delphi
- MIT multi-host agent workflows with self-update and self-modification examples (discovery only; review persistence and rollback): https://github.com/XAIHT/Tlamatini
- Apache-2.0 multi-host coding-agent Skills for release publishing, testing, cloud workflows, code review, and Skill authoring: https://github.com/OEvortex/vtx-coding-agent
- MIT research and reasoning Skills for hypothesis generation, knowledge acquisition, verification, quality gates, and automation (review self-evolution and side effects): https://github.com/yakeworld/Synthos
- MIT index of 5,400+ OpenClaw skills (discovery only; verify host assumptions, scripts, licenses, and permissions before WorkBuddy adaptation): https://github.com/VoltAgent/awesome-openclaw-skills
- MIT cross-platform Agent Skill Manager for WorkBuddy and other domestic Agent products (review install scopes and sync paths before copying or removing): https://github.com/yxdwind/agent-skill-manager
- MIT read-only Skill Auditor for usage traces and security hazards (deletion suggestions require human confirmation): https://github.com/dx2020z/skill-auditor
- MIT Skill Packer for frontmatter validation and platform-ready ZIPs (inspect the archive and license before publishing): https://github.com/dx2020z/skill-packer
- Apache-2.0 Google Workspace CLI with Agent Skills for Gmail, Drive, and Calendar (not an official Google product; review OAuth and credentials): https://github.com/googleworkspace/cli
- MIT Obsidian Agent Skills for Markdown, Bases, JSON Canvas, and Obsidian CLI (review Vault path and file-write scope): https://github.com/kepano/obsidian-skills
- MIT GitHub-maintained Awesome Copilot collection of agents, instructions, prompts, and Skills (review host and repository permissions): https://github.com/github/awesome-copilot
- MIT TypeScript-focused Agent Skills collection by Matt Pocock (review exact paths and command assumptions): https://github.com/mattpocock/skills
- MIT Wshobson development workflow agents and Skills (review scripts, dependencies, and host configuration): https://github.com/wshobson/agents
- MIT Vercel Skills CLI for listing, previewing, and prompt-only use across compatible hosts (not the WorkBuddy desktop ZIP importer): https://github.com/vercel-labs/skills
- MIT Microsoft-maintained Agent Skills for developer and cloud workflows (review tenant, credentials, and host assumptions): https://github.com/microsoft/skills
- CC BY 4.0 Microsoft Learn-derived Agent Skills (retain attribution; review docs, versions, tenant, and subscription): https://github.com/MicrosoftDocs/Agent-Skills
- MIT Microsoft Azure Agent plugin with Skills and MCP configurations (review tenant, subscription, identity, writes, and billing): https://github.com/microsoft/azure-skills
- Apache-2.0 Google Gemini API, SDK, and model/agent Skills (review keys, access, quota, retention, and code assumptions): https://github.com/google-gemini/gemini-skills
- Apache-2.0 Elastic-maintained Agent Skills (review cluster, API keys, index permissions, and telemetry): https://github.com/elastic/agent-skills
- Apache-2.0 n8n automation Skills (review credentials, webhooks, side effects, and deployment scope): https://github.com/n8n-io/skills
- Apache-2.0 NVIDIA-maintained AI and GPU Agent Skills (review toolchains, model downloads, GPU access, and cost): https://github.com/NVIDIA/skills
- Apache-2.0 AWS Agent Toolkit for AWS workflows (review least-privilege account, region, write, and billing boundaries): https://github.com/aws/agent-toolkit-for-aws
- Apache-2.0 Cloudflare-maintained platform Agent Skills (review account permissions and deployment effects): https://github.com/cloudflare/skills
- MIT Supabase Agent Skills for database and platform workflows (review project, migration, and secret boundaries): https://github.com/supabase/agent-skills
- Apache-2.0 dbt Agent Skills for analytics engineering and testing (review warehouse credentials, SQL, and writes): https://github.com/dbt-labs/dbt-agent-skills
- Apache-2.0 official MongoDB Agent Skills (review OAuth, database targets, and writes): https://github.com/mongodb/agent-skills
- MIT Microsoft Waza CLI for Agent Skill evaluation (review binaries, fixtures, and model/API cost): https://github.com/microsoft/waza
- MIT Everything Claude Code agent-harness reference for skills, memory, security, research-first development, and multi-host workflows (inspect host-specific hooks): https://github.com/affaan-m/ECC
- MIT Marketing Skills for CRO, copywriting, SEO, analytics, and growth engineering (review claims, customer data, and publishing permissions): https://github.com/coreyhaines31/marketingskills
- Apache-2.0 defensive cybersecurity Skills mapped to common security frameworks (authorize scans, credentials, and targets): https://github.com/mukul975/Anthropic-Cybersecurity-Skills
- MIT portable Humanizer Skill for reducing AI-sounding phrasing (editorial aid; verify claims and required disclosure): https://github.com/blader/humanizer
- Apache-2.0 collection of agents, RAG apps, MCP projects, and agent_skills (discovery only; review dependencies, credentials, and data flows): https://github.com/Shubhamsaboo/awesome-llm-apps
- MIT Planning with Files workflow for long-running tasks, recovery, and completion gates (keep WorkBuddy paths explicit): https://github.com/OthmanAdi/planning-with-files
- MIT PM Skills collection for discovery, strategy, execution, launch, and growth (review business-sensitive inputs): https://github.com/phuryn/pm-skills
- Apache-2.0 Agent Skills specification and documentation repository (check format and portability assumptions): https://github.com/agentskills/agentskills
- Apache-2.0 EverMind SkillCorpus pipeline for provenance, license/safety gates, retrieval, evaluation, and export (architecture reference; review sources separately): https://github.com/EverMind-AI/SkillCorpus
- MIT Keep the Why repo-native convention and Agent Skill for preserving decisions and rejected alternatives (adapt the habit; inspect host-specific layout): https://github.com/oliver-zehentleitner/keep-the-why
- Apache-2.0 self-hosted iFlytek SkillHub registry (review Docker, auth, and package provenance): https://github.com/iflytek/skillhub
- Apache-2.0 Hugging Face Agent Skills for Hub workflows (review tokens, permissions, artifacts, and compute cost): https://github.com/huggingface/skills
- Community task-oriented guide: https://github.com/AlephAITech/WorkBuddyGuide
- Trail of Bits security Skills (CC-BY-SA-4.0; inspect attribution and authorization): https://github.com/trailofbits/skills
- NVIDIA SkillSpector security scanner (Apache-2.0; review aid, not a trust guarantee): https://github.com/NVIDIA/SkillSpector
- Gentle AI cross-agent configuration reference (MIT): https://github.com/Gentleman-Programming/gentle-ai
- HTML Anything local agentic HTML workflow (Apache-2.0; inspect sandbox and publishing boundaries): https://github.com/nexu-io/html-anything
- Chinese Skill navigation reference with no formal repository license: https://github.com/anbeime/skill
- MIT WorkBuddy account migration reference (diagnose first; modifies local sessions, memory, and connector data): https://github.com/xiaoliuzhuan666/workbuddy-account-migrate
- MIT WorkBuddy-compatible self-media compliance review (local evidence by default; optional real-time channels may require credentials, network access, and paid quota): https://github.com/JuneYaooo/self-media-compliance-review
- MIT WorkBuddy-compatible QianLV UI design reference (inspect bundled assets and scaffold scripts before reuse): https://github.com/CodeGanHaoZ/gks-QianLV-uiSkill
- Apache-2.0 WorkBuddy-compatible CodeDrobe theme workflow (review CDP scope and restart approvals): https://github.com/CodeDrobe/skills
- Apache-2.0 CodeDrobe runtime for WorkBuddy theming (pin the runtime and review CDP/app-path permissions): https://github.com/CodeDrobe/core
- MIT WorkBuddy-compatible ontology-driven development reference (inspect runtime, examples, and model/API configuration): https://github.com/sharptoolbox/ontology-driven-dev
- MIT focused WorkBuddy AppBuilder reference for requirements, domain modeling, and app generation (review generated-artifact and model/API boundaries): https://github.com/sharptoolbox/WorkBuddy-AppBuilderSkill
- MIT confirmation-gated audio-first image-story video workflow for Codex and WorkBuddy (inspect media paths and approvals): https://github.com/aaronyi97/image-story-video-wizard
- MIT Agent Skills package explicitly listing WorkBuddy support for distilling reusable thinking models (review research and privacy boundaries): https://github.com/alchaincyf/nuwa-skill
- WorkBuddyGuide online reading site: https://workbuddy.homes
- MIT community WorkBuddy usage and troubleshooting reference: https://github.com/Neo5093/workbuddy-guide
- MIT WorkBuddy task-flow and rollback-boundary reference: https://github.com/maning636/workbuddy-runbook
- MIT Chinese WorkBuddy Skills collection for research, memory, context, stock analysis, and learning (review WorkBuddy frontmatter and Codex compatibility boundary): https://github.com/dingdong905/workbuddy-skills
- Chinese WorkBuddy practice guide (verify freshness and license before reuse): https://github.com/opcspace/WorkBuddy-Guide
- MIT WorkBuddy beginner starter: https://github.com/sunyet-01/WorkBuddy-Starter
- MIT WorkBuddy-compatible QA testing reference: https://github.com/genapohub/qa-testing-guide
- Apache-2.0 WorkBuddy-compatible Chinese WeChat article workflow (review credentials and publishing permissions): https://github.com/aiworkskills/wechat-article-skills
- MIT WorkBuddy/Codex-compatible WeChat article-to-Markdown/Feishu workflow (review rights, image uploads, OAuth, and destination permissions): https://github.com/bonboruyau-dev/wechat2knowledge
- MIT Chinese writing workflow for official-media, social-media, and brand formats (verify claims, attribution, style boundaries, and platform rules): https://github.com/ZOORO-NEW/qianjin-writer
- MIT Chinese trend-research workflow for topic discovery, lifecycle, platform adaptation, and compliance prompts (verify freshness and sources): https://github.com/ZOORO-NEW/qianjin-trending-hunter
- MIT agentic Skill catalog and control plane (research reference): https://github.com/sickn33/agentic-awesome-skills
- MIT Chinese job-search and development Skills with a documented WorkBuddy bridge (review browser and resume-data actions): https://github.com/Hisn00w/ASu-skills
- Auto-updated Agent Skills index with no repository license declaration (navigation only): https://github.com/linny006/awesome-agent-skills
- MIT portable Skill discovery and evidence workflow (methodology only; not an automatic installer): https://github.com/CodeSigils/skill-discovery
- Apache-2.0 Agent toolchain that exports actions and playbooks as Skills, MCP servers, HTTP services, or binaries (separate builder; inspect generated files): https://github.com/team4u/actiondock
- Apache-2.0 Agent-ready web interoperability and provenance reference covering llms.txt, Agent Skills, MCP, and A2A (research reference): https://github.com/dkharlanau/agent-ready-web-profile
- MIT unofficial reversible WorkBuddy UI-theme reference using local CDP injection (review debug-port exposure and rollback): https://github.com/itcastWsy/workbuddy-skin
- MIT unofficial WorkBuddy Skin Studio using loopback CDP (high-risk reference; inspect scripts, debug port, and rollback): https://github.com/cdredfox/workbuddy-skin-studio
- MIT cross-harness Skill and extension manager with preview and source tracking (separate tool): https://github.com/vanillagreencom/kendex
- MIT zero-dependency local Skill usage and context-cost report (disable/prune actions require exact-target review): https://github.com/bborok1234/skillmon
- CC0 WorkBuddy resource index: https://github.com/semlinker/awesome-workbuddy
- Chinese AI coding and WorkBuddy tutorial: https://github.com/KimYx0207/AI-Coding-Guide-Zh
- MIT WorkBuddy-compatible e-commerce visual-copywriting Skill (inspect bundled assets and scripts before reuse): https://github.com/feichanggege/ecommerce-visual-copywriting-skill
- MIT Chinese multi-domain expert-role collection (inspect attribution and bundled assets before reuse): https://github.com/jnMetaCode/agency-agents-zh
- MIT offline-first WorkBuddy usage analytics dashboard (inspect credential boundaries): https://github.com/clancy-feng/workbuddy-usage-status
- MIT local token-usage analysis CLI with WorkBuddy support (review trace paths and redaction): https://github.com/YuLaiZ/token-usage
- MIT WorkBuddy Windows OCR/click/scheduled-task check-in automation (high risk; review platform terms and every script before running): https://github.com/KristoffersonLee/workbuddy-checkin
- MIT WorkBuddy usage-analysis tool (reads local traces, database, and usage logs; share only redacted reports): https://github.com/Elisabeth15501/agent-analytics-report
- MIT early local MCP task-pool reference for WorkBuddy/Codex collaboration (not production-hardened): https://github.com/somo-ui/workbuddy-codex-hub-mcp
- MIT WorkBuddy local chat recovery/cleanup reference (high-risk database and session-file operations; review backups and confirmations): https://github.com/sutupikk-cyber/workbuddy-chat-manager
- MIT local AI-asset portability tool with WorkBuddy support (auto-memory and account data are not migrated): https://github.com/muzishuiji/mnemoport
- MIT WorkBuddy cross-device sync and handoff reference (review database, identity, path, and credential boundaries): https://github.com/jamesting-eng/workbuddy-skills
- MIT WorkBuddy document and web-page-to-Markdown skill (inspect SSRF and script boundaries): https://github.com/stwhwing/markitdown-skill
- Clean-room WorkBuddy architecture course: https://github.com/adongwanai/learn-workbuddy
- WorkBuddy Harness v2 architecture and evaluation reference (no formal LICENSE file; research only): https://github.com/zhuang-HE/workbuddy-harness
- MIT WorkBuddy community skills collection: https://github.com/yinqd3/workbuddy-skills
- Apache-2.0 Huawei Cloud DevKit with explicit WorkBuddy support and safety checks (research reference): https://github.com/huaweicloud/huaweicloud-devkit
- Tencent CloudBase AI Toolkit with explicit WorkBuddy support for backend, Skills, MCP, and deployment workflows (research reference): https://github.com/TencentCloudBase/CloudBase-AI-Toolkit
- TencentDB Agent Memory with WorkBuddy integration and reusable memory assets; README shows MIT but GitHub reports no formal license (research reference only): https://github.com/TencentCloud/TencentDB-Agent-Memory
- Apache-2.0 Qwen multimodal Skill + optional MCP plugins (research reference): https://github.com/QwenLM/Qwen-MM-Plugins
- MIT macOS Skill manager with explicit WorkBuddy/CodeBuddy support (research reference): https://github.com/crossoverJie/SkillDeck
- MIT zero-dependency Skill/MCP governance CLI with WorkBuddy directory support (research reference): https://github.com/GrubbyLee/skill-manager
- MIT cross-platform Skill manager with WorkBuddy sync support (research reference): https://github.com/qufei1993/skills-hub
- MIT bilingual WorkBuddy port and workflow conventions: https://github.com/mrzhangguoguo/oh-my-workbuddy
- WorkBuddy Harness blue book (research reference; no license declaration): https://github.com/zjp1997720/zhijian-ai-bluebook-workbuddy-harness
- WorkBuddy Harness blue book online edition: https://zjp1997720.github.io/zhijian-ai-bluebook-workbuddy-harness/
- Tencent WorkBuddy benchmark: https://github.com/Tencent/workbuddy-bench
- MIT external draw.io XML generation and validation workflow for WorkBuddy-compatible agents (reference only): https://github.com/bruc3van/bruce-drawio
- MIT Tencent EdgeOne website Prompt/Skill gallery (external navigation only; review each entry separately): https://github.com/TencentEdgeOne/awesome-website-prompts-and-skills
- GitSkills public Skill dataset research: https://arxiv.org/abs/2608.10906
- SWE-Skills-Bench requirement-driven evaluation: https://arxiv.org/abs/2603.15401
- SkillsBench paired Skill/no-Skill evaluation: https://arxiv.org/abs/2602.12670
- Agent Skill Evaluation and Evolution survey: https://arxiv.org/abs/2606.11435
- Community Awesome WorkBuddy index (navigation only): https://github.com/staruhub/awesome-workbuddy
- WorkBuddy multi-agent skill groups (inspect attribution and licenses): https://github.com/darker2016/workbuddy-skill-groups
- WorkBuddy public-market archive (offline study; respect original rights): https://github.com/infometa/workbuddyskills
- Community WorkBuddy skills collection (navigation only; inspect license and scripts): https://github.com/bitcjm/workbuddy-skills
- WorkBuddy ChatCut MCP integration (MIT; inspect OAuth, credentials, and media-data scope): https://github.com/chonpszhou/workbuddy-chatcut-mcp
- Discussions for usage examples and reviewed-package feedback: https://github.com/sandbaseai/workbuddy-skill/discussions
- English support and troubleshooting: https://github.com/sandbaseai/workbuddy-skill/blob/main/SUPPORT.md
- Chinese support and troubleshooting: https://github.com/sandbaseai/workbuddy-skill/blob/main/SUPPORT.zh-CN.md
- Documentation, usability, validation, and existing-package feedback (the public catalog is frozen): https://github.com/sandbaseai/workbuddy-skill/issues/new?template=feature.yml
- English quickstart: https://github.com/sandbaseai/workbuddy-skill/blob/main/docs/quickstart.md
- Chinese quickstart: https://github.com/sandbaseai/workbuddy-skill/blob/main/docs/quickstart.zh-CN.md
- Skill lifecycle (update, disable, remove, and rollback): https://github.com/sandbaseai/workbuddy-skill/blob/main/docs/quickstart.md#update-disable-or-remove-a-skill
- Skill troubleshooting (import, connector, schema, and authorization failures): https://github.com/sandbaseai/workbuddy-skill/blob/main/docs/quickstart.md#troubleshooting
- Cross-platform release verifier: https://github.com/sandbaseai/workbuddy-skill/blob/main/scripts/verify_release.py
- English resource map: https://github.com/sandbaseai/workbuddy-skill/blob/main/docs/resources.md
- Chinese resource map: https://github.com/sandbaseai/workbuddy-skill/blob/main/docs/resources.zh-CN.md
- English task-based Starter Packs: https://github.com/sandbaseai/workbuddy-skill/blob/main/docs/starter-packs.md
- Chinese task-based Starter Packs: https://github.com/sandbaseai/workbuddy-skill/blob/main/docs/starter-packs.zh-CN.md
- Catalog reading guide: https://github.com/sandbaseai/workbuddy-skill/blob/main/docs/catalog-guide.md
- Chinese catalog reading guide: https://github.com/sandbaseai/workbuddy-skill/blob/main/docs/catalog-guide.zh-CN.md
- Official WorkBuddy quickstart: https://www.workbuddy.ai/docs/zh/workbuddy/Quickstart
- Official WorkBuddy overview and reading order: https://www.workbuddy.ai/docs/zh/
- Official Agents quickstart (tasks, parallel work, conversations, and results): https://www.workbuddy.ai/docs/zh/ide/User-guide/Agent-Mode/Quickstart
- Official WorkBuddy Memory guide: https://www.workbuddy.ai/docs/zh/workbuddy/From-Beginner-to-Expert-Guide/Function-Description/Memory
- Official WorkBuddy macOS installation guide: https://www.workbuddy.ai/docs/workbuddy/From-Beginner-to-Expert-Guide/Installation-Mac-Guide
- Official WorkBuddy Windows installation guide: https://www.workbuddy.ai/docs/workbuddy/From-Beginner-to-Expert-Guide/Installation-Win-Guide
- Official WorkBuddy model configuration guide: https://www.workbuddy.ai/docs/workbuddy/From-Beginner-to-Expert-Guide/Function-Description/Model
- Official WorkBuddy Assistant remote-control guide: https://www.workbuddy.ai/docs/workbuddy/From-Beginner-to-Expert-Guide/Function-Description/Assistant
- Official DingTalk remote-control integration guide: https://www.workbuddy.ai/docs/zh/workbuddy/Platform-Integration/Dingtalk-Guide
- Official Yuanbao remote-control integration guide: https://www.workbuddy.ai/docs/zh/workbuddy/Platform-Integration/YuanBaoPai-Guide
- Official WorkBuddy Skill Marketplace guide: https://www.workbuddy.ai/docs/workbuddy/From-Beginner-to-Expert-Guide/Function-Description/Skills-Market
- Official WorkBuddy Privacy Policy: https://www.workbuddy.ai/document/privacy-policy
- Official WorkBuddy Service Agreement: https://www.workbuddy.ai/document/term
- Official WorkBuddy Acceptable Use Policy: https://www.workbuddy.ai/document/acceptable-use-policy
- Official WorkBuddy Permission Modes: https://www.workbuddy.ai/docs/workbuddy/From-Beginner-to-Expert-Guide/Function-Description/Permission-Modes
- Official WorkBuddy Task Bar: https://www.workbuddy.ai/docs/workbuddy/From-Beginner-to-Expert-Guide/Function-Description/Task-Bar
- Official large repositories and monorepos guide: https://www.workbuddy.ai/docs/cli/large-codebases
- Official WorkBuddy tips and tricks: https://www.workbuddy.ai/docs/zh/workbuddy/From-Beginner-to-Expert-Guide/Efficient-Tips
- Official WorkBuddy FAQ: https://www.workbuddy.ai/docs/zh/workbuddy/From-Beginner-to-Expert-Guide/FQA
- Official WorkBuddy release notes (SkillHub, plugins, connectors, models, and platform updates): https://www.workbuddy.ai/docs/zh/ide/release-notes/release-notes
- Vendor-neutral Agent Skills specification: https://agentskills.io/specification
- Official Skills tutorial: https://www.workbuddy.ai/docs/zh/workbuddy/From-Beginner-to-Expert-Guide/Practice-Cases/Create-Skills
- Official MCP guide: https://www.workbuddy.ai/docs/zh/workbuddy/From-Beginner-to-Expert-Guide/Function-Description/MCP-Guide
- Official Automation guide: https://www.workbuddy.ai/docs/workbuddy/From-Beginner-to-Expert-Guide/Function-Description/Automation-Guide
- Chinese official Automation guide: https://www.workbuddy.ai/docs/zh/workbuddy/From-Beginner-to-Expert-Guide/Function-Description/Automation-Guide
- English official quickstart: https://www.workbuddy.ai/docs/workbuddy/Quickstart
- English official Skills tutorial: https://www.workbuddy.ai/docs/workbuddy/From-Beginner-to-Expert-Guide/Practice-Cases/Create-Skills
- English official MCP guide: https://www.workbuddy.ai/docs/workbuddy/From-Beginner-to-Expert-Guide/Function-Description/MCP-Guide
- Official Skill Marketplace: https://www.workbuddy.ai/docs/zh/workbuddy/From-Beginner-to-Expert-Guide/Function-Description/Skills-Market
- Official Explore: https://www.workbuddy.ai/docs/zh/workbuddy/From-Beginner-to-Expert-Guide/Function-Description/Explore
- Chinese official WorkBuddy Task Bar: https://www.workbuddy.ai/docs/zh/workbuddy/From-Beginner-to-Expert-Guide/Function-Description/Task-Bar
- English official Skill Marketplace: https://www.workbuddy.ai/docs/workbuddy/From-Beginner-to-Expert-Guide/Function-Description/Skills-Market
- English official Explore: https://www.workbuddy.ai/docs/workbuddy/From-Beginner-to-Expert-Guide/Function-Description/Explore
- Official Open Platform Skill guide: https://open.workbuddy.cn/en/docs/skill
- Official Open Platform overview: https://open.workbuddy.cn/en/docs/what-is-open-platform
- Official Open Platform onboarding: https://open.workbuddy.cn/en/docs/onboarding
- Official Buddy App guide: https://open.workbuddy.cn/en/docs/buddy-app
- Official Expert guide: https://open.workbuddy.cn/en/docs/expert
- Official Expert Team guide: https://open.workbuddy.cn/en/docs/expert-team
- Official Open Platform Connector guide: https://open.workbuddy.cn/en/docs/connector
- Official Open API reference: https://open.workbuddy.cn/en/docs/openapi
- Official Third-Party App guide: https://open.workbuddy.cn/en/docs/third-party-app
- Chinese Open Platform Skill guide: https://open.workbuddy.cn/zh/docs/skill
- Chinese Open Platform overview: https://open.workbuddy.cn/docs/what-is-open-platform
- Chinese Open Platform onboarding: https://open.workbuddy.cn/docs/onboarding
- Chinese Buddy App guide: https://open.workbuddy.cn/docs/buddy-app
- Chinese Expert guide: https://open.workbuddy.cn/docs/expert
- Chinese Expert Team guide: https://open.workbuddy.cn/docs/expert-team
- Chinese Open Platform Connector guide: https://open.workbuddy.cn/docs/connector
- Chinese Open API reference: https://open.workbuddy.cn/docs/openapi
- Chinese Third-Party App guide: https://open.workbuddy.cn/docs/third-party-app
- GitHub CLI gh skill manual (preview): https://cli.github.com/manual/gh_skill
- Adaptation guide: https://github.com/sandbaseai/workbuddy-skill/blob/main/docs/adapting-skills.md
- Chinese adaptation guide: https://github.com/sandbaseai/workbuddy-skill/blob/main/docs/adapting-skills.zh-CN.md
- Copy-ready use cases: https://github.com/sandbaseai/workbuddy-skill/blob/main/docs/use-cases.md

## Safety and provenance

The Atlas stores metadata, immutable source links, compatibility signals, and conservative static review signals. A clean signal is not a security guarantee. Review the original license, instructions, bundled resources, scripts, network behavior, and permissions before adapting or installing any Skill.

For a copy-ready pre-publication content-review workflow, use the “Pre-publication content review” section in `docs/use-cases.md`. Keep it local and read-only by default; real-time platform searches require explicit confirmation of tools, credentials, network access, and cost.

## Install a reviewed package

Use a package's `asset` value from `packages.json` in the following command, then verify the ZIP before importing it into WorkBuddy under **Experts · Skills · Connectors → Skills → Add Skill**:

```bash
mkdir -p workbuddy-download
gh release download --repo sandbaseai/workbuddy-skill \
  --pattern '<asset>' --pattern SHA256SUMS \
  --dir workbuddy-download --clobber
cd workbuddy-download && sha256sum --check SHA256SUMS --ignore-missing
```

For the complete package-set check, run `python3 scripts/verify_release.py
workbuddy-download` from a local repository checkout; it also rejects extra
WorkBuddy ZIPs that are absent from `SHA256SUMS`.

The human-readable package page exposes the exact `download_command` for every reviewed package. Do not install a catalog-only record as if it were a reviewed package; inspect its immutable source first.

## Local query

Clone the repository and run:

```bash
python3 scripts/query_catalog.py research --security no-static-flags --source-context primary-looking --min-score 80 --unique --sort score --limit 10
```

For the same review-oriented defaults, use `python3 scripts/query_catalog.py research --high-signal --limit 10`.

To preview a repository's current SKILL.md files without writing the catalog, use
`python3 scripts/crawl_github_skills.py --dry-run --repository owner/name --repository-only --target 100`.
For a repeatable source list, use
`--repository-file config/upstream-skill-sources.txt`; blank lines and `#` comments are ignored.

Use `--json` for machine-readable results. Search terms are combined, so every term must match the repository, path, inferred skill name, metadata, or full catalog ID. Atlas filters are available for WorkBuddy package availability, WorkBuddy status, static review state, category, minimum score, unique blob SHA, and result ordering. The local CLI also accepts `--category research`, `--package-status reviewed`, or `--package-status catalog-only`; reviewed results include the direct Release ZIP URL and a copy-ready GitHub CLI command, exposed as `workbuddy_package_url` and `workbuddy_download_command` in JSON. Every JSON result also exposes its inferred or curated Atlas category as `workbuddy_category`.

## Adapt a reviewed result

```bash
python3 scripts/adapt_skill.py \
  --catalog-id 'github:owner/repository:path/to/SKILL.md' \
  --display-name-zh '中文名称' \
  --display-name-en 'English name' \
  --description-zh '中文说明' \
  --description-en 'English description' \
  --author 'Original author; WorkBuddy adapter' \
  --source-license 'MIT'
```

The adapter does not execute downloaded content. It records provenance, applies size and path limits, statically scans instructions and referenced scripts, and packages referenced resources from the same immutable commit.