Open protocol and portable .skill format for AI skills.
Install once. Point your AI at the work.
Agents create, inspect, hand off, and dry-run skills — you review and approve releases.
skillerr.com ·
dot-skill/skillerr ·
npm i -g skillerr
Plain markdown “skills” and chat exports break down: every model re-interprets prose, context dies across tools, and there is no integrity story before something runs.
.skill is a sealed, inspectable package — typed I/O, workflow, pinned knowledge, redacted provenance, digests, optional mint. Skillerr is the open protocol; skillerr is the reference CLI your agent uses.
The mark is the Dotling — the living . in .skill.
npm i -g skillerrThen paste this into Cursor, ChatGPT, Claude, Codex, or any agent with shell tools:
Install skillerr if needed. Set SKILL_HOST to your host id. From this conversation,
create a portable .skill with a redacted journey and exact sections I approved
(secrets as {{refs}}). Checkpoint for handoff, or compile --approve --mint when
release-complete. Do not invent filler. Show status and the output path.
More agent prompts and docs: dot-skill/skillerr · skillerr.com
Inspect TrustView (digests/seals) before run. Declared host/model fields are self-reported provenance — not cryptographic proof of authorship. Reference mint HMAC in the repo is development-only.