Skillerr
Decentralized seals. Digest-pinned. Inspect before run.
A decentralized registry of sealed skills with TrustView. Digest-pinned installs, declared capabilities, inspect before run.
npx @skillerr/add skillerr/find-skillsWhy sealed
Markdown is input. The product is a sealed package.
Public skill dumps hand you a floating file. Skillerr mints a digest-pinned .skill, shows TrustView before run, and lets you verify the pin against the registry.
01
Discover
Browse providers and community skills with declared permissions.
02
Trust
Skill Score, provenance, and honest Anchored / Not anchored.
03
Install sealed
npx @skillerr/add writes a .skill ZIP, not a bare SKILL.md.
04
Inspect & verify
Open TrustView, check the digest, confirm it matches the registry.
See TrustView on find-skills · Why sealed & verify · Browse sealed catalog
From Anthropic
Official · AnthropicFrom Vercel
Official · VercelFrom Supabase
Official · SupabaseFrom the community
Submitted by users- skillerr/find-skills
Discover sealed skills on skillerr.com by intent, capability, or keyword: the meta-skill agents install first.
12.8k installs - shadcn/shadcn
Add and compose shadcn/ui components the idiomatic way.
9.1k installs - firecrawl/firecrawl
Scrape and crawl the web into clean markdown for agent consumption.
6.3k installs - skillerr/create-skill
Author a complete SkillContract with typed I/O, workflow steps, and human approve gates.
5.1k installs - obra/systematic-debugging
Root-cause debugging with hypotheses, evidence, and minimal fixes: no shotgun patches.
4.6k installs - skillerr/inspect-trust
Inspect a .skill TrustView (digest, capabilities, mint status) before you run anything.
4.2k installs - skillerr/seal-skill
Convert a SKILL.md or chat journey into a sealed, digest-pinned .skill package.
3.9k installs - mattpocock/tdd
Test-driven development workflow for agents: red, green, refactor with discipline.
3.2k installs
Live activity
live- anonymousdownloadedskillerr/score-signals0s ago
- @maya_chendownloadedskillerr/find-skills1h ago
- @dev_jordandownloadedskillerr/inspect-trust1h ago
- @nova_labsdownloadedskillerr/seal-skill1h ago
- @tess_wdownloadedmattpocock/tdd1h ago
- @priya_sdownloadedskillerr/create-skill1h ago
- @bytecraftdownloadedobra/systematic-debugging1h ago
- @nina_vdownloadedsupabase/supabase-postgres-best-practices2h ago
- @alex_kimdownloadedskillerr/publish-skill2h ago
- @sam_okorodownloadedskillerr/capability-audit2h ago
- @kit_andersdownloadedshadcn/shadcn2h ago
- @lena_bergdownloadedskillerr/handoff-continuity2h ago
Trust principles
Digest-pinned
Every install resolves to a content hash. What you inspect is what you run.
Capabilities declared
fs, net, shell: skills state what they can touch before you grant trust.
Transparency log
Publishes and installs land in an append-only log you can audit anytime.