Tools
Every Skill, MCP server, CLI and plugin in the directory. Filter by type, profession or tag, or search by keyword.
Profession: Full-stack / Indie +4
Pick one or more: resources that suit any of them are listed.
- Full-stack / Indie522
- DevOps / SRE209
- ML / AI162
- Data analysis152
- Operations233
- Frontend447
- Backend408
- Mobile & desktop96
- Game development29
- QA101
- Data engineering81
- GIS / Surveying12
- Security71
- Product / Projects246
- Design239
- Architecture / Civil19
- Hardware engineering17
- Marketing & sales306
- Admin / Office238
- Writing / Content229
- Video / Media187
- Customer support65
- HR88
- Legal94
- Finance / Accounting142
- Investing / Analysis104
- Teachers / Students212
- Research / Academia255
- Healthcare19
- Translation / L10n84
- Founders280
Tags
Pick one or more: only resources carrying all of them are listed.
- python104
- data-analysis89
- cloud80
- web-development73
- anthropic69
- terminal69
- dev-workflow65
- deployment58
- code-review57
- claude-code55
- subagents55
- social-media49
- database48
- llm47
- backend46
- copywriting43
- agent-development42
- codex42
- task-management42
- code-quality41
There are more. Type to find them.
1,011 items

Engineering
PluginEveryday software engineering plugin: standup summaries, code review, structured debugging, architecture decision records, incident response and pre-deploy checks. Official from Anthropic, built mainly for Cowork but also usable in Claude Code.
anthropics/knowledge-work-plugins32.2k
Bio Research
PluginFor preclinical life sciences research: bundles MCP connectors for PubMed, bioRxiv, ChEMBL, Open Targets, ClinicalTrials and more, plus analysis skills for single-cell RNA QC, scvi-tools and nf-core pipelines. Official from Anthropic; some data sources need accounts.
anthropics/knowledge-work-plugins32.1k
Process Doc
SkillTurns processes that live in people's heads into SOPs: flowcharts, a RACI matrix (responsible, accountable, consulted, informed), detailed steps and exceptions, ready for handoffs, audits and onboarding. Describe the process or paste existing notes. Official from Anthropic.
anthropics/knowledge-work-plugins32.1k
Grafana k6
CLIGrafana's load and performance testing tool: write test scripts in JavaScript covering HTTP, WebSocket, gRPC and browsers, set threshold assertions and run them in CI.
grafana/k632k
Context Mode
PluginProcesses large tool outputs in a sandbox and returns only the results to context; logs session events in SQLite with FTS5 so work continues after context compaction. Supports Claude Code, Codex and a dozen other clients. ELv2 licensed.
mksglu/context-mode32k
Marketing
PluginMarketing plugin: drafts blog, social, email and landing page content, plans campaigns, reviews copy against your brand voice, and produces competitive briefs, SEO audits and channel performance reports. Official from Anthropic.
anthropics/knowledge-work-plugins32k
campaign-plan
SkillGenerates a full campaign plan from a marketing goal: objectives, audience, key messages, channel strategy, a week-by-week content calendar and success metrics, plus budget allocation if you give a budget. Suited to launches, acquisition and brand campaigns. From Anthropic's marketing plugin.
anthropics/knowledge-work-plugins32k
status-report
SkillGenerates project or team status reports with an executive summary, overall red/yellow/green status, key metrics, risks and open decisions, turning scattered project tool activity into a report for leadership. From Anthropic's official operations plugin.
anthropics/knowledge-work-plugins32k
Cowork Plugin Management
PluginGuides you through creating a plugin from scratch or customizing an existing plugin's connectors, skills and settings for your team's tools and workflows, producing an installable plugin package. Maintained by Anthropic; runs in Cowork mode in the Claude desktop app.
anthropics/knowledge-work-plugins32k
Data
PluginData analysis plugin: writes queries in each SQL dialect, explores datasets, runs statistical tests, charts with Python and builds interactive HTML dashboards, then reviews methods and conclusions before sharing. Connects to Snowflake, BigQuery and other warehouses. Official from Anthropic.
anthropics/knowledge-work-plugins31.9k
Context Mode
MCPRuns commands and scripts in a sandbox, indexes large tool outputs for full-text search and returns only results to the context window to save tokens; after compaction it can retrieve edits and task state. Supports Claude Code, Codex and more. Needs Node.js 22.5+.
mksglu/context-mode31.9k
persona-customer-support
SkillSupport persona skill for the gws CLI: triages support-labeled Gmail, turns emails into tasks, logs tickets in Google Sheets, escalates urgent issues to Chat and books follow-ups. Needs gws authorized via your own Google Cloud project plus its companion skills; not officially supported by Google.
googleworkspace/cli31.8k
Google Workspace CLI Skills
SkillSkill collection for the gws CLI: one skill per Workspace API plus recipes for common Gmail, Drive, Docs, Calendar and Sheets workflows. Install gws and authorize your Google account first; not an officially supported Google product.
googleworkspace/cli31.8k
gws (Google Workspace CLI)
CLIOne CLI for Gmail, Drive, Calendar, Docs, Sheets and other Google Workspace APIs, with commands generated from the Discovery service, JSON output and 100+ Agent Skills. From the googleworkspace org but not officially supported; needs OAuth via a Google Cloud project.
googleworkspace/cli31.8k
gws-forms
SkillBuilt-in gws CLI skill: lets agents create Google Forms, add or edit questions in bulk, publish and read responses, handy for sign-ups, surveys and class feedback. Needs gws with a Google account authorized via your own Google Cloud project; not an officially supported Google product.
googleworkspace/cli31.8k
Lighthouse
CLIOfficial web auditing tool from the Google Chrome team: score pages on performance, accessibility, best practices and SEO from the command line and get HTML, JSON or CSV reports, so agents can iterate on a page. Needs Node 22+ and Chrome.
GoogleChrome/lighthouse30.9k
Serena
MCPGives coding agents IDE-like symbol-level code search, editing and refactoring via language servers (LSP) across 40+ languages, with cross-session memory. An optional paid JetBrains plugin backend adds breakpoint debugging. Install with uv first.
oraios/serena30.7k
OpenTofu
CLIOpen-source infrastructure-as-code tool and community fork of Terraform: describe cloud resources declaratively, run plan to generate an execution plan, then apply changes, creating resources in parallel by dependency graph. Community-governed, MPL-2.0.
opentofu/opentofu30.6k
OpenCLI
CLITurns websites and Electron desktop apps into deterministic CLIs, with built-in adapters for Reddit, Hacker News, Bilibili, Zhihu, Xiaohongshu and more, and lets agents operate any page through your signed-in Chrome. Needs its companion browser extension.
jackwener/OpenCLI30.6k
Helm
CLIKubernetes package manager that packages, installs, upgrades and rolls back releases with Chart templates, including ready-made charts from Artifact Hub. A CNCF graduated project; needs a kubeconfig to reach your cluster.
helm/helm30.4k
Gitleaks
CLIDetects hardcoded passwords, API keys and tokens in Git repos, files and stdin, and can run as a pre-commit hook or GitHub Action to block leaks. Now feature-frozen with security fixes only, as the author has moved on to Betterleaks.
gitleaks/gitleaks30.3k
video-use
SkillBrowser Use's video editing skill: put raw footage in a folder and chat with Claude Code or Codex to get a finished cut, with filler words and pauses removed, color grading, burned-in subtitles, animations and per-cut self-checks. Needs ffmpeg, Python (uv) and an ElevenLabs API key.
browser-use/video-use29.6k
Repomix
CLIPacks a whole repository into a single AI-friendly file (XML, Markdown, JSON or plain text) with per-file token counts, automatically excludes files containing secrets, and can compress code with Tree-sitter or pack remote GitHub repos directly.
yamadashy/repomix29k
Claude HUD
PluginStatus line plugin for Claude Code showing the model, Git branch, context usage and rate limits live below the input, optionally with running tools, active subagents and todo progress.
jarrodwatts/claude-hud28.8k