Trusted Claude Skills

Curated catalog of Claude Code (and Claude Desktop) Agent Skills with trust ratings. Tiers: Anthropic / First-party / Vetted-third-party / Personal-use / Audit-before-use.

PublicOwned byJacob ColeSource:wikihub.md/@jacobcole/trusted-claude-skills/index· Imported from WikiHub @jacobcole/trusted-claude-skills/index

Cozy Compact Dense

You're viewing this list as a guest. Sign in to comment or star items.Sign in

Reviews changed code for reuse, quality, and efficiency.

Recurring task runner; calls ScheduleWakeup / cron.

Manages scheduled remote agents (routines).

Anthropic SDK helper — prompt caching, model migration.

Guided feature development with codebase awareness.

Bootstraps a new CLAUDE.md from a codebase scan.

Produces non-generic UI code; avoids generic AI aesthetics.

Generates a PRD for a new feature.

Converts PRD → Ralph autonomous-agent JSON.

Mac App Store submission walkthrough.

Edits ~/.claude/settings.json and hooks.

Scans transcripts and proposes allow-list rules.

Customizes ~/.claude/keybindings.json.

Branch security review.

PR review skill.

Anthropic's own public skill repo. Default trust.github.com

Every skill scanned for malware, prompt injection, and credential theft.www.skillsdirectory.com

7,000+ AI-evaluated skills; S-rank (9.0+) skills are the conservative bets.www.skillhub.club

Verified-tested-safe skill registry; positions against the 13% malicious baseline.github.com

Security-first vetting. Re-audit anything before piping to a tool with shell access.lobehub.com

Community aggregator from GitHub. Useful for discovery; treat each skill as untrusted until read.skillsmp.com

140+ open-source skills, but no security audit before listing.claudeskills.info