brewtoolsagent-togglelisted

# Agent Toggle > **Disable/enable individual agents** via the native `permissions.deny` mechanism in `settings.json`. A `Agent(<name>)` deny entry removes the subagent from the model context. State survives plugin updates — **no SessionStart reapply hook needed**. See also: `/brewtools:skill-toggle` for skills. <instructions> ## Robustness Rules | Rule | Applies | |------|---------| | Every Bash call ends with `&& echo "OK ..." \|\| echo "FAILED ..."` | ALL | | Never use `Write`/`Edit` on any `settings.json` — use Bash + Node helper | ALL | | This skill mutates **agents only** (`permissions.deny`) — never skills | ALL | | `Agent(name)` matches the BARE name — strip any `plugin:` prefix the user passes | ALL | | Atomic writes via lockfile + temp + rename (handled by helper) | P2 | Paths (substitute literally in Bash): - Helper: `$CLAUDE_PLUGIN_ROOT/skills/_shared/toggle/deny.mjs` - Catalog: `$CLAUDE_PLUGIN_ROOT/skills/_shared/toggle/cache.mjs` (`enumeratePlugins`) - Settings per scope: - `global` -> `~/.claude/settings.json` - `project` -> `<cwd>/.claude/settings.json` - `local` -> `<cwd>/.claude/settings.local.json` ## How deny works `permissions.deny: ["Agent(<bareName>)"]` natively disables a subagent (plugin or project-custom), removing it from the model's tool context. The name is the BARE agent name (`Agent(ssh-admin)`, NOT `Agent(brewtools:ssh-admin)`). Deny-first precedence across scopes. Takes effect on next session or `/reload-plugins`. > **COLLISION WA