Commands

What Claude touched this session: reads, edits (+/− lines), searches per file, main vs subagent, with flags for blind edits and repeated reads.

Git at a glance and in reach: a branch/changes readout, a /git pane to diff, stage, revert and commit, and Conventional Commit messages written by haiku.

/handoff writes a structured handoff (goal, state, decisions, files, next steps, gotchas, how to verify) to .claude/handoffs/; /handoff load hands it to a later session.

Project memory across sessions: every few turns haiku writes what was done, decided and left open to .claude/journal.md, and new conversations start with the latest entries.

Whole-tree time travel: a git snapshot of every non-ignored file before each turn, with diff, restore and /undo, never touching HEAD, the index or the stash.

Ask before guessing: when a prompt looks ambiguous, Claude asks up to three clarifying questions first. Off until /clarify on.

Inline #tokens that attach live context to a prompt: #diff, #staged, #log:20, #tree, #todo, #errors, #file:path:10-40, #url:https://...

Switchable working styles (teach, terse, pair, review, tdd, explore, ship, or your own) applied to every prompt you send.

Standing context you control: pin notes and files once, and every prompt you send carries them.

Starts every conversation oriented: one context block with the repo, branch, layout, stack and its build/test/lint/run commands, README summary and recent commits.

Every prompt you write, kept across sessions and projects: /history searches them and puts one back in the prompt box.

Rough idea in, precise prompt out: /polish or a trailing ?? rewrites your prompt into the input box for you to edit and send.

Park a half-written prompt, ask something else, and pop it back into the box later.

Text expansion in the prompt box: type ;plan and a space, get the full instruction.

Blocks catastrophic shell commands (rm -rf ~, curl | sh, force-push to main, DROP DATABASE, format C:) and asks before risky ones (git reset --hard, rm -rf, terraform apply), in Bash and PowerShell.

Keeps work off protected branches: no commit, merge, rebase, cherry-pick, am or push on main/master/release/* (or an explicit push to one), with an optional auto-branch for commits.

Stops accidental truncation of real code: no "// ... rest of the code" placeholders over real code, and a question before a Write or Edit throws away most of a file.

Supply-chain check before installs: npm/pnpm/yarn/bun/npx, pip/uv/poetry/pipx, cargo and gem packages are looked up first; nonexistent names are refused, brand-new, unpopular, deprecated or lookalike ones need your yes.

Protected files and a project boundary: Claude can't edit .env, keys or .git, asks before touching lockfiles and migrations, and asks before writing outside the project.

Notices the commands you keep approving by hand and offers a one-click standing allow rule (never for risky or interpreter-wide commands).

"Look, don't touch": /readonly on blocks edits and every command that is not provably read-only, and tells Claude to investigate and explain instead.

Keeps credentials out of your code, your shell commands and Claude's context: blocks hard-coded secrets in edits, asks before commands with literal keys, redacts secrets from tool results and flags them in prompts.

Commit messages in your repository's own style (learned from its last 30 subjects) and pull request descriptions from the branch, drafted by haiku; nothing is committed or pushed without a press.

Opt-in: questions about the codebase (where is, what does, how does, which file, find) go to a read-only haiku scout subagent first; its answer comes with Ask Claude anyway and Good enough, and the tokens saved.