| name | linkdrop-x-0x-kaize-awesome-claude-code-repos |
| description | Pointer to hesreallyhim/awesome-claude-code — the canonical community curation of Claude Code skills, agents, hooks, slash-commands, plugins, statuslines, and MCP servers (39,800+ stars, actively maintained). Use to discover community solutions before building new Claude Code tooling from scratch. |
| source | https://x.com/0x_kaize/status/2045580955759309131 → https://github.com/hesreallyhim/awesome-claude-code |
| allowed-tools | Read, Write, Edit, Bash, WebFetch, Glob, Grep |
linkdrop-x-0x-kaize — Awesome Claude Code
Overview
hesreallyhim/awesome-claude-code (39,800+ stars, updated daily) is the canonical community-curated index of Claude Code resources. Skills, agents, hooks, slash-commands, plugins, statuslines, workflows — if a community solution exists for a CC workflow problem, it's cataloged here.
The author's thesis (@0x_kaize): most CC users never discover this list and end up rebuilding patterns that already exist. This skill exists so the agent consults the community layer before composing a new solution.
Local-first principle: this pairs with toolkit-scout (which scans your 1,100+ local skills). Toolkit-scout covers the local layer; this skill covers the community layer one hop beyond that. Run toolkit-scout FIRST.
The canonical list
Real categories (from the live README)
| Category | What's in it |
|---|
| Agent Skills | Topic-organized skill definitions ready to drop into ~/.claude/skills/ |
| Workflows & Knowledge Guides | General / Teams / Ralph Wiggum sub-sections |
| Tooling | General, IDE Integrations, Usage Monitors, Orchestrators, Config Managers |
| Status Lines | Custom statusline implementations (rate pace, token burn, etc.) |
| Hooks | Pre/post-tool, Stop, UserPromptSubmit, SessionStart hooks |
| Slash-Commands | General, Version Control & Git, Code Analysis & Testing, Context Loading & Priming, Documentation & Changelogs, CI / Deployment, Project & Task Management, Miscellaneous |
| CLAUDE.md Files | Language-Specific / Domain-Specific / Project Scaffolding & MCP |
| Alternative Clients | Non-terminal CC clients |
| Official Documentation | Anthropic's canonical docs |
The README is ~45 KB and updated faster than any static mirror. Always pull it live rather than cache a snapshot here.
Workflow — community-before-build
Step 1 — Local layer first (always):
cat ~/.claude/skills/toolkit-scout/SKILL.md
bash ~/.claude/scripts/route.sh "<your task>" --top=5
Step 2 — If local miss, pull the awesome-claude-code README:
gh api repos/hesreallyhim/awesome-claude-code/readme --jq '.content' | base64 -d > /tmp/acc-readme.md
grep -A 20 "^## Hooks" /tmp/acc-readme.md
grep -A 30 "^## Slash-Commands" /tmp/acc-readme.md
grep -A 20 "^### IDE Integrations" /tmp/acc-readme.md
Step 3 — Evaluate a candidate entry (trust gate):
gh repo view <owner>/<repo> --json pushedAt,stargazerCount,licenseInfo,description
Step 4 — Security pre-install scan (mandatory for any community repo):
mkdir -p ~/sandbox/<repo> && cd ~/sandbox/<repo>
gh repo clone <owner>/<repo> .
bash ~/.claude/scripts/phase3/check-secrets.sh .
bash ~/.claude/scripts/phase3/check-gbrain-ban.sh .
find . -name '*.sh' -o -name 'hooks.json' | xargs grep -l 'curl\|wget\|nc ' || echo "no network hooks"
Step 5 — Integrate additively, never overwrite:
Copy individual patterns into ~/.claude/skills/linkdrop-community-<source>-<item>/ — never install whole packs globally without trust review. Whole-pack installs bypass our security gates.
Step 6 — Log what you imported:
Add an entry to docs/knowledge-base/wiki/02-workflow-and-dx/linkdrop-x-0x-kaize-awesome-claude-code-repos.md (or a new KB file) under "Community tools evaluated" — what you took, from where, with trust-gate notes.
When to use
- Starting any non-trivial Claude Code build and
toolkit-scout returned thin local results.
- Evaluating "is there an existing hook / slash-command / plugin for X?" before writing one.
- Onboarding a new engineer to the CC ecosystem after they've read local CLAUDE.md.
- Auditing a
.claude/ setup — compare against community-canonical patterns.
- Building a curation list or "starter kit" for a new sub-team.
Do NOT use when:
- The problem is domain-specific (deal origination, PB / CapIQ data pipelines) — awesome-claude-code is horizontal, not domain.
- You need vetted-for-security tools — community repos require your own trust review every time.
Gotchas
- Trust is NOT transitive. A repo linked from awesome-claude-code is curated for popularity or utility, not for safety. Always run the security scan.
- Plugin rot is faster than skill rot. Prefer extracting individual patterns to installing whole packs.
- GitHub API rate limit:
gh search repos is 30/min unauthenticated, 5000/hr authenticated. If you're authenticated via gh auth login, you're fine.
- License compliance per
~/.claude/rules/dependency-management.md: MIT / Apache-2.0 / BSD / ISC accepted; GPL / AGPL / SSPL flagged for legal review before use.
- Overlap with local: many community skills duplicate patterns already in your 1,100+ local inventory. Running
toolkit-scout FIRST is not optional.
Setup / Prerequisites
gh CLI authenticated (gh auth status).
- Local sandbox dir (
~/sandbox/ — not tracked in main git).
~/.claude/scripts/phase3/ validators present: check-secrets.sh, check-gbrain-ban.sh, check-skill-density.sh.
- Familiarity with the local skill format: frontmatter + Overview + When to use + Workflow + density-script compliance.
Output contract
When invoked with a specific category (e.g., "find a hook for auto-commit"), return:
category: hooks
candidates:
- repo: <owner>/<repo>
stars: N
last_push: YYYY-MM-DD
license: MIT
path_in_repo: <relative/path/to/pattern>
trust_gate: pass | review | reject
recommendation: <which to import, with one-line rationale>
security_scan:
banned_terms: pass | fail
secrets: pass | fail
network_hooks: none | <list>
Source
Cross-references
toolkit-scout — MUST run before this skill (local-first principle).
saraev-cc-claude-code-plugin-discovery — plugin install UX.
anthropic-claude-code-best-practices-linkdrop — Anthropic's official guidance.
~/.claude/rules/dependency-management.md — license + vulnerability rules.
~/.claude/scripts/phase3/check-secrets.sh — mandatory pre-import secret scan.
docs/knowledge-base/wiki/02-workflow-and-dx/INDEX.md — local workflow KB.
Related KB entry
docs/knowledge-base/wiki/02-workflow-and-dx/linkdrop-x-0x-kaize-awesome-claude-code-repos.md — narrative version, trust-gate protocol, past community imports.
Link-drop Additions
(Reserved — append idempotently when same author drops new threads.)