You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
This repository was archived by the owner on Jul 30, 2026. It is now read-only.
The README/SKILL lead agent-first, which causes recurring "why install this as a skill / why not just a plugin?" confusion (it came up repeatedly).
Reframe the umbrella README + skill README to lead with the audiences:
CLI for you — a .env-replacement you run at your terminal (a human doesn't need the "skill" framing).
Skill for your agent, everywhere — the portable AgentSkills artifact (Claude Code, Agent SDK, claude.ai, OpenClaw).
Plugin for Claude Code — the enforced superset (guard hook + MCP + /kdbx:*) that bundles the skill.
Key facts to surface: plugins are Claude-Code-only, so the skill is the portable core; the agent's real value is run/check (use secrets without seeing them). Resolves the "why a skill" + "why not plugin-only" questions for the next reader.
The README/SKILL lead agent-first, which causes recurring "why install this as a skill / why not just a plugin?" confusion (it came up repeatedly).
Reframe the umbrella README + skill README to lead with the audiences:
.env-replacement you run at your terminal (a human doesn't need the "skill" framing)./kdbx:*) that bundles the skill.Key facts to surface: plugins are Claude-Code-only, so the skill is the portable core; the agent's real value is
run/check(use secrets without seeing them). Resolves the "why a skill" + "why not plugin-only" questions for the next reader.