Claude Code component directories compared on the same 6 criteria

2 columns

Every figure below was read off that company's own page; each source URL and the date it was read are listed at the foot of this page. SkillWorks is the publisher of this table and is not one of the 2 compared. It scores these rather than competing with them.

subagents.shawesome-claude-code

Items listed40 subagents

Printed on the site's own header beside its install total: "40 subagents · 323 installs". The smallest catalogue of the four by two orders of magnitude, and it is a deliberate scope rather than an incomplete crawl.

Not published

The repository publishes no total for how many resources the list carries — its headline figures are repository figures: 52.3k stars, 4.6k forks, 1,530 commits. Read on the repository's own page on 14 August 2026.

What it listsSubagents only

"Subagents are reusable capabilities for Claude Code. They extend Claude's abilities with specialized knowledge and workflows for specific tasks." No skills, plugins, commands or MCP servers.

Skills, agents, status lines, tooling, plugins — and prose

"A delectable showcase of top tier skills, ambidextrous agents, scintillating status lines, top notch developer tooling, and also we have plugins." Its table of contents also carries categories no installer covers: Documentation, Knowledge & Learning; Alternative Clients; Usage & Cost Observability.

How something gets listedHost it on GitHub; no approval

"Sharing your subagent is simple - no approval process required": put the file in .claude/agents/, agents/ or the repository root, share the install command, and "your subagent will automatically appear on the leaderboard as people install it".

Issue form, gated on age or stars

"Any resource that is recommended must either: (i) Be at least 14 days old … AND show signs of active development … OR (ii) Have at least 100 stars." One at a time, via the web issue form only — "Do not open a PR" — and "resources that fail these criteria will be closed automatically". The only published maturity bar of the four.

Whether anything checks the itemNot published

The documentation publishes the file format a subagent must follow — name required, description and tools optional — but no statement that anything validates a listed file against it, and no per-listing health signal. Read on subagents.sh's own docs page on 14 August 2026.

A human reads it; the bot does not

Stated plainly by the maintainer: "The bot is a mechanical form-validator, it does not review the quality of the recommended resource", and "there is no formal submission/review process at the moment … Recommendations are reviewed in a best-effort way". The check is editorial, not mechanical, and the page says so rather than implying otherwise.

How you install from itnpx @augmnt-sh/subagents add owner/repo/name

"This downloads the subagent to .claude/agents/ in your project directory. Once installed, Claude Code will automatically discover and use the subagent when relevant." `list`, `remove` and `update` are documented alongside it.

Not published

There is no installer, CLI or manifest — the list is links and descriptions, and installation is whatever each linked project documents for itself. Read on the repository's own page on 14 August 2026.

How the list is orderedInstall count, from CLI telemetry

"The leaderboard is sorted by install count. We use anonymous telemetry from CLI installations to track popularity. No personal information is collected." The only ordering of the four that measures the artifact rather than its repository.

Not published

Entries are grouped by category — Start Here, From Anthropic, Security, Agent Orchestration and so on — with no stated ordering inside a category and no sort control. Read on the repository's own page on 14 August 2026.

“Not published” means that company does not state the figure on the page cited at the foot of this page. It is not a claim that no such figure exists. A company that publishes no limit most likely has none, but “most likely” is not a fact, and this table prints what was read rather than what was inferred.

Whether anything checks the item. whether the directory itself validates that a listed skill, subagent or plugin parses and loads, as distinct from checking that its author is popular

subagents.sh vs awesome-claude-code — items listed?

subagents.sh: 40 subagents. Printed on the site's own header beside its install total: "40 subagents · 323 installs". The smallest catalogue of the four by two orders of magnitude, and it is a deliberate scope rather than an incomplete crawl. awesome-claude-code does not state it on github.com. The repository publishes no total for how many resources the list carries — its headline figures are repository figures: 52.3k stars, 4.6k forks, 1,530 commits. Read on the repository's own page on 14 August 2026. Read from subagents.sh and github.com on 2026-09-05.

subagents.sh vs awesome-claude-code — what it lists?

subagents.sh: Subagents only. "Subagents are reusable capabilities for Claude Code. They extend Claude's abilities with specialized knowledge and workflows for specific tasks." No skills, plugins, commands or MCP servers. awesome-claude-code: Skills, agents, status lines, tooling, plugins — and prose. "A delectable showcase of top tier skills, ambidextrous agents, scintillating status lines, top notch developer tooling, and also we have plugins." Its table of contents also carries categories no installer covers: Documentation, Knowledge & Learning; Alternative Clients; Usage & Cost Observability. Read from subagents.sh and github.com on 2026-09-05.

subagents.sh vs awesome-claude-code — how something gets listed?

subagents.sh: Host it on GitHub; no approval. "Sharing your subagent is simple - no approval process required": put the file in .claude/agents/, agents/ or the repository root, share the install command, and "your subagent will automatically appear on the leaderboard as people install it". awesome-claude-code: Issue form, gated on age or stars. "Any resource that is recommended must either: (i) Be at least 14 days old … AND show signs of active development … OR (ii) Have at least 100 stars." One at a time, via the web issue form only — "Do not open a PR" — and "resources that fail these criteria will be closed automatically". The only published maturity bar of the four. Read from subagents.sh and github.com on 2026-09-05.

When should I pick subagents.sh, and when awesome-claude-code?

subagents.sh — You want subagents specifically and you want the one-line install more than you want the catalogue. It is the only one of the four with a dedicated CLI for the artifact — `npx @augmnt-sh/subagents add owner/repo/name` writes straight into .claude/agents/ and there are `list`, `remove` and `update` commands beside it — and the only one whose ordering is a real usage signal rather than a stars count: "The leaderboard is sorted by install count. We use anonymous telemetry from CLI installations to track popularity.". awesome-claude-code — You want a human's judgement rather than a search result, and you are browsing rather than installing. It is the only one of the four where a person decided each entry was worth your attention — "A hand-picked collection … with an emphasis on code quality, security, and originality" — and the only one that covers things no installer can install: field guides, essays, status lines, alternative clients, cost-observability tooling. At 52.3k stars it is also the single most-read document in this ecosystem, which makes it the best place to find out what exists at all.

When each of these is the right pick

SkillWorks does not sell any of these and has nothing to gain from which one you choose. It scores them. Each case below is the one that option genuinely wins.

subagents.sh. You want subagents specifically and you want the one-line install more than you want the catalogue. It is the only one of the four with a dedicated CLI for the artifact — `npx @augmnt-sh/subagents add owner/repo/name` writes straight into .claude/agents/ and there are `list`, `remove` and `update` commands beside it — and the only one whose ordering is a real usage signal rather than a stars count: "The leaderboard is sorted by install count. We use anonymous telemetry from CLI installations to track popularity."
awesome-claude-code. You want a human's judgement rather than a search result, and you are browsing rather than installing. It is the only one of the four where a person decided each entry was worth your attention — "A hand-picked collection … with an emphasis on code quality, security, and originality" — and the only one that covers things no installer can install: field guides, essays, status lines, alternative clients, cost-observability tooling. At 52.3k stars it is also the single most-read document in this ecosystem, which makes it the best place to find out what exists at all.

Other Claude Code component directories compared head to head

3 more

What was checked, and when

4 sources

Prices and limits move, and a page quoting a stale one is worth less than a page quoting none. Every source is printed here so a reader can check it against the day they are reading this, and a job re-fetches each of these URLs on a schedule and fails when a figure above no longer appears on the page it came from.

subagents.sh, subagents.sh/. Items listed. Read 2026-09-05.
subagents.sh, subagents.sh/docs. What it lists, How something gets listed, Whether anything checks the item, How you install from it, How the list is ordered. Read 2026-09-05.
awesome-claude-code, github.com/hesreallyhim/awesome-claude-code. Items listed, What it lists, How you install from it, How the list is ordered. Read 2026-09-05.
awesome-claude-code, github.com/hesreallyhim/awesome-claude-code/blob/main/CONTRIBUTING.md. How something gets listed, Whether anything checks the item. Read 2026-09-05.

Last verified 2026-09-05.