Record, search, and prune project learnings in `.claude/learnings.jsonl`. Learnings accumulate across sessions and are typed, confidence-scored, and prunable.

Full release pipeline: tests → coverage → review gate → version bump → CHANGELOG → commit → PR. Each step is a hard gate — failure stops the pipeline and reports clearly.

Execute TODO.md steps via unattended Claude Code sessions

Generate detailed content briefs for blog posts with target keywords,

Generates a morning briefing: what ran overnight, what it cost, what to do next.

Start, stop, or manage Clade autonomous coding loops remotely

Get session reports, cost breakdowns, and blocker details from Clade

Check Clade loop progress, worker status, costs, and recent commits

Generate a PreToolUse or PostToolUse hook skeleton from a mature CLAUDE.md rule

The user just saw a set of options you offered (numbered issues with A/B/C sub-options, or a recommendation block with alternatives). They don't want to re-read them — they want the one YOU recommended to be executed now.

Squash-merges a PR, deletes the remote branch, and cleans up the local branch. Parallel to OpenClaw's `/merge-pr`.

Check your API usage quota (auto-detects Minimax vs Claude Code)

Searches the web for latest Claude model information, compares benchmarks and pricing, then updates the model selection guide.

Act as a project orchestrator, not a code writer. Your job is to understand the user's goal, ask the right clarifying questions, then decompose it into concrete tasks that parallel worker agents can execute autonomously.

The user hit `esc` during a long generation and wants a heartbeat: are you stuck, still progressing, or done? Answer in ≤3 lines, then continue if fine.

Toggle the statusline quota pace indicator (mode/theme)

One command to launch autonomous coding sessions. Morning briefings, targeted goals, overnight runs, cross-project patrols — all from `/start`.