feat(plugin): add setup skill — guided watchmen install across agents#124
Merged
Conversation
A thin orchestrator skill (`/watchmen:setup` in Claude Code, `/skills setup` or `$setup` in Codex) that walks a user through installing and wiring watchmen: detects uv + existing install, installs the CLI, hands off the interactive `watchmen init` wizard and the agent-only `/plugin` steps, wires the statusline, and verifies with `watchmen doctor`. Cross-agent: covers both Claude Code and Codex. Never touches credentials — defers all provider auth to the init wizard. - plugin/skills/setup/SKILL.md (Claude variant, with allowed-tools) - plugin-codex/skills/setup/SKILL.md (Codex variant: $-invocation, AGENTS.md, no statusline, no allowed-tools — matching the brief skill's conventions) - Bump both plugin manifests 0.1.7 -> 0.1.8 (the #123 CI guard requires a bump when plugin/skills changes; keeps the two harnesses in version lockstep). Note: shipping in the plugin means a brand-new user can't invoke it until they install the plugin (the thing it sets up) — it's primarily for wiring a second agent, repairing, or verifying. A standalone-install README line can follow. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Adds a 'Guided setup skill' note to the Plugins section: how to invoke /watchmen:setup (or $setup in Codex) once installed, and how to curl the SKILL.md into ~/.claude/skills/ standalone so a brand-new machine can run it before any plugin exists (resolving the in-plugin chicken-and-egg). Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Why
Setting up watchmen is a multi-step flow (uv tool install →
watchmen initwizard → agent/pluginsteps → statusline → verify), and the one step the CLI genuinely can't do is the agent's own/plugininstall — that's interactive agent UI. A skill running inside the agent can orchestrate the whole thing and adapt per agent.What it does
/watchmen:setup(Claude Code) //skills setupor$setup(Codex) is a thin orchestrator:uv+ any existing install, branches accordingly (won't reinstall over a working setup)git clone+uv sync && uv tool install --editable .)watchmen initwizard (waits for the user — never drives provider auth or touches keys)/plugincommands for whichever agent(s) the user has, and wires the statusline (Claude Code)watchmen doctorand walks the ✓/✗/! linesCross-agent by design: covers both Claude Code and Codex. It's an orchestrator over existing commands, not a reimplementation of the wizard.
Files
plugin/skills/setup/SKILL.md— Claude variant (withallowed-tools)plugin-codex/skills/setup/SKILL.md— Codex variant ($-invocation, AGENTS.md, no statusline, noallowed-tools), matching the existingbriefskill's conventions0.1.7 → 0.1.8(the fix(plugin): bump version to 0.1.7 so /plugin picks up changes + guard it #123 guard requires a bump whenplugin/skills/changes)Known limitation
Shipping in the plugin means a brand-new user can't invoke it until they install the plugin — i.e. the thing it sets up. So it's primarily for wiring a second agent, repairing, verifying, or updating. A standalone-install README line (curl the SKILL.md into
~/.claude/skills/) can follow if we want it usable pre-install.Testing