This VoltAgent repository is a collection of about 160 Claude Code subagents: Markdown files that give Claude a specialist role, a set of tools and instructions for one kind of work. They are grouped into ten categories, from core development (API design, backend, frontend, mobile) and language specialists (TypeScript, Python, Rust, Swift, SQL and many more) to infrastructure, quality and security, data and AI, developer experience, specialised domains, business and product, multi-agent orchestration, and research.
Each category installs as its own Claude Code plugin, so you can add the language specialists without the business analysts. You can also copy single files into ~/.claude/agents or a project's .claude/agents folder, use an interactive installer script, or install an agent-installer subagent that lets Claude browse and install the rest for you.
It has about 26,000 stars, 42 contributors and an MIT licence. Like any prompt collection, quality varies from file to file, and a subagent is only as good as the model and context behind it; read one before relying on it.
- Repository: github.com/VoltAgent/awesome-claude-code-subagents
- Licence: MIT (MIT License)
- Language: Shell. Stars: 26.1K. Forks: 2,997. Last push: Oct 11, 2026.
- Scan: safe, Oct 26, 2026, commit 408d3c9
Who it is for
Claude Code users who want ready-made specialist subagents for their stack instead of writing their own, and people learning how to write good subagent definitions by example.
Getting started
1. Add the repository as a Claude Code plugin marketplace
claude plugin marketplace add VoltAgent/awesome-claude-code-subagents2. Install a category, for example the language specialists
claude plugin install voltagent-lang3. Or browse and install single agents with the interactive script
git clone https://github.com/VoltAgent/awesome-claude-code-subagents.git && cd awesome-claude-code-subagents && ./install-agents.shOther categories follow the same pattern, such as voltagent-infra and voltagent-core-dev; the README lists them all. The orchestration agents in voltagent-meta work best with other categories installed.
Safety scan
We cloned VoltAgent/awesome-claude-code-subagents at commit 408d3c9 on Oct 26, 2026 and ran the checks described on the GitHub Tools page: credential patterns, decode-and-execute code, install-time scripts, committed binaries, risky CI workflows, every host the code talks to, known vulnerabilities in pinned dependencies, and project hygiene. A person read every hit. This is what we found.
- No secrets, no suspicious code patterns, no committed binaries and no hooks across 198 files and about 41,700 lines, almost all Markdown. A skill-specific pass for invisible Unicode and prompt-injection phrases found nothing, and the plugin marketplace ships no hooks or MCP configuration.
- install-agents.sh (584 lines, no sudo) lists categories through api.github.com and downloads agent files from this repository's raw.githubusercontent.com path into ~/.claude/agents or .claude/agents. The README's standalone option downloads the script to disk before running it rather than piping it to a shell.
- 105 of the agents ask for Read, Write, Edit, Bash, Glob and Grep, so they can edit files and run shell commands when Claude delegates to them. Claude Code's permission settings still apply, but check an agent's tools line before installing it.
- One workflow with no third-party actions. Licence and contributing guide present; no security policy or Dependabot. The README carries sponsor listings, and the contribution rules reject agents written to promote a product.
What the scanner counted
| Check | Result |
|---|---|
| Secrets | None found. |
| Suspicious code | None found. |
| Install-time code | 1 installer script |
| Committed binaries | None. |
| CI workflows | 1 workflow. None use pull_request_target. No third-party actions. |
| Network hosts | 2 distinct hosts referenced from source; most often raw.githubusercontent.com, api.github.com. No URLs to bare IP addresses. |
| Known vulnerabilities | No lockfile to check: dependencies are declared as ranges, so what gets installed is whatever is current on the day. |
| Project hygiene | Has licence file, contributing guide. Missing security policy, automated dependency updates, CodeQL. |
| OpenSSF Scorecard | Not scored: the project is not in Scorecard's weekly index. |
The raw findings
Every hit the scanner wrote out, with a link to the exact line at the scanned commit. Secrets candidates are redacted.
Installer scripts (1)
- install-agents.sh, 584 lines; talks to api.github.com, raw.githubusercontent.com
By the numbers
| Stars | 26.1K |
|---|---|
| Forks | 2,997 |
| Contributors | 42 |
| Commits | 527 |
| Open issues | 2 |
| Open pull requests | 2 |
| Releases | 0 |
| Latest release | none tagged |
| Licence | MIT |
| Main language | Shell |
| Project age | 1 year |
| Last push | Oct 11, 2026 |
| Tracked files | 198 |
| Lines of code | 41.7K |
| Checkout size | 1 MB |
Lines by language: Markdown 40.5K, Shell 679, JSON 410, YAML 82.
Questions
Is Awesome Claude Code Subagents free?
Yes. The collection is MIT-licensed and free to install and modify. You need Claude Code, under Anthropic's own pricing, to use the subagents.
What is a Claude Code subagent?
A subagent is a Markdown file with a name, a description, a list of allowed tools and a system prompt. Claude Code can hand a task to it, and it works in its own context window, which keeps the main conversation focused.
Should I install every category?
No. Each installed agent adds its description to what Claude considers when delegating, so install the categories that match your work and remove agents you never use.
This post is part of GitHub Tools, where every repository is cloned and scanned before it is written up. The scan is a snapshot of one commit on one day; the repository has moved on since, so check it before you install.