From autodev
Execute a trivial task inline — no planning, no subagents, just do it
How this command is triggered — by the user, by Claude, or both
Slash command
/autodev:autodev-fast <task description>This command is limited to the following tools:
The summary Claude sees in its command listing — used to decide when to auto-load this command
<objective> Execute a trivial task directly in the current context without spawning subagents or generating TASK.md files. For tasks too small to justify planning overhead: typo fixes, config changes, small refactors, simple additions. </objective> <process> <step name="parse_task" Parse `$ARGUMENTS` for the task description. If empty, use AskUserQuestion: </step> <step name="scope_check" **Before doing anything, verify this is actually trivial.** A task is trivial if it can be completed in: - ≤ 3 file edits - ≤ 1 minute of work - No new dependencies or architecture changes - No resea...
<step name="parse_task"
Parse $ARGUMENTS for the task description.
If empty, use AskUserQuestion:
What's the quick fix? (one sentence)
<step name="scope_check" Before doing anything, verify this is actually trivial.
A task is trivial if it can be completed in:
If the task seems non-trivial:
This looks like it needs planning. Use `/autodev-plan` instead.
And stop.
<step name="execute_inline" Do the work directly:
No TASK.md. Just do it.
<step name="log_to_state"
If .autodev/STATE.md exists, append to the "Quick Tasks" section.
If the section doesn't exist, skip silently.
<step name="done" Report completion:
✅ Done: {what was changed}
Files: {list of changed files}
No next-step suggestions. No workflow routing. Just done.
- NEVER spawn a Task/subagent — this runs inline - NEVER create TASK.md or SUMMARY.md files - NEVER run research or plan-checking - If the task takes more than 3 file edits, STOP and redirect to `/autodev-plan`<success_criteria>
npx claudepluginhub mthanhlm/autodev --plugin autodev