From mypos-legal-copilot
STUB -- references the published Devil's advocate review skill. This stub exists so the plugin can ship without bundling the skill body. The actual skill body is published in the `anthropic-skills` plugin (or wherever the team has published it). Use this skill via the `triage-reviewer` subagent which loads the published version.
How this skill is triggered — by the user, by Claude, or both
Slash command
/mypos-legal-copilot:devils-advocate-reviewThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
This file is a placeholder. The plugin does NOT bundle the Devil's advocate skill body -- it lives in the published Anthropic skills plugin (or the myPOS-internal skills marketplace, depending on where it was published).
This file is a placeholder. The plugin does NOT bundle the Devil's advocate skill body -- it lives in the published Anthropic skills plugin (or the myPOS-internal skills marketplace, depending on where it was published).
When /triage calls the triage-reviewer subagent (agents/triage-reviewer.md), that subagent calls Skill(skill: "devils-advocate-review") which Claude Code resolves to the published skill regardless of which plugin published it. The plugin user must have access to that skill in their environment.
If at any point we want the plugin to ship its own copy of the Devil's advocate skill (e.g., to pin a known-good version, or to operate without the published one):
SKILL.md with the full skill body (frontmatter + instructions).name: to keep devils-advocate-review (so callers don't change).description: to whatever the bundled version describes..claude-plugin/plugin.json./setup-copilot runs a smoke test that confirms the skill resolves. If it doesn't, the user is told to either:
anthropic-skills)Confirm with the legal copilot maintainers (Atanas / Emil) where the Devil's advocate skill is published and whether plugin users will reliably have it available. If yes, this stub is fine. If publishing is uncertain, replace with a bundled copy.
Provides behavioral guidelines to reduce common LLM coding mistakes, focusing on simplicity, surgical changes, assumption surfacing, and verifiable success criteria.
Searches, retrieves, and installs Agent Skills from prompts.chat registry using MCP tools like search_skills and get_skill. Activates for finding skills, browsing catalogs, or extending Claude.
Creates, edits, and optimizes skills for Claude Code, including drafting, evaluating with test prompts, iterating on performance, and improving skill descriptions for better triggering accuracy.
npx claudepluginhub mypostech/mps-legal-legalcopilot-ai