/loopx
loopx goal-mode setup (NOT Claude Code's built-in /goal). `/loopx <task>` sets up a goal + writes .claude/loop.md; drive the loop with native /loop. bare /loopx = arm; off | status.
$ npx -y skills add loopx-project/loopx --agent claude-codeHow it fires
How this command gets triggered: by you, by Claude, or both.
- Fires itselfClaude auto-loads it when your prompt matches the work.
- You can call itInvoke it directly when you want it.
- Slash command
/loopx
Context preview
What this command does when you run it.
loopx goal-mode setup (NOT Claude Code's built-in /goal). `/loopx <task>` sets up a goal + writes .claude/loop.md; drive the loop with native /loop. bare /loopx = arm; off | status.
Command definition
loopx.mddescription: loopx goal-mode setup (NOT Claude Code's built-in /goal). `/loopx <task>` sets up a goal + writes .claude/loop.md; drive the loop with native /loop. bare /loopx = arm; off | status. argument-hint: <task to do> | (no args = arm) | off | status allowed-tools: Bash(python3:*)
Run the loopx setup helper and read its output:
!`python3 "${CLAUDE_PLUGIN_ROOT}/scripts/goalmode_cmd.py" $ARGUMENTS`
The output is loopx control-plane SETUP / status info — it is the COMPLETE, user-facing result. Show it to the user VERBATIM and STOP. Do NOT do any work, claim todos, write files, or loop here: the WORK runs when the user types Claude Code's native `/loop` (which executes `.claude/loop.md`). Do not summarize a multi-line `status` detail block into one line.
NOTE: this is `/loopx` (loopx control-plane SETUP), NOT a runtime and NOT Claude Code's built-in `/goal`. The run loop is native `/loop`; loopx provides the deterministic `should_run` protocol.
A control plane with a durable state kernel for long-horizon agents and teams. Keep work moving and improving across sessions, with less human attention.
Repo: loopx-project/loopx