Stats
Actions
Tags
From skeft-qa
Toggle the AskUserQuestion guard for the current managed /goal loop. Turn it off when you want the loop to be able to ask you a question; turn it back on to resume autonomous protection. Use when you want to enable or disable being asked questions during an autonomous dev loop.
How this skill is triggered — by the user, by Claude, or both
Slash command
/skeft-qa:goal-guard <on|off><on|off>This skill is limited to the following tools:
The summary Claude sees in its skill listing — used to decide when to auto-load this skill
<!-- disable-model-invocation: true is USER-ONLY by deliberate policy (spec 11 refined): an
Toggle the PreToolUse(AskUserQuestion) guard by flipping question_guard in this
session's marker. Default is on in goal mode (questions are intercepted + redirected).
Turn it off when you genuinely want to be asked; on to resume autonomous protection.
The session id is passed as an argument (the substituted ${CLAUDE_SESSION_ID}), so the
toggle does not depend on an env var being exported to the shell:
Run: !`uv run --no-sync python "${CLAUDE_PLUGIN_ROOT}/scripts/goal_guard_toggle.py" $ARGUMENTS ${CLAUDE_SESSION_ID} 2>/dev/null`
Then confirm the new state to the user (question_guard=on|off). If the line above is empty
or reads [shell command execution disabled by policy], tell the user shell execution is
disabled (the toggle could not run). If it reports no marker for this session, the session is
not a managed loop — there is nothing to toggle.
npx claudepluginhub netrxn/sk_eft_hawking --plugin skeft-qaProvides a checklist for code reviews covering functionality, security, performance, maintainability, tests, and quality. Use for pull requests, audits, team standards, and developer training.