/goal

Drive the milestone named in `$ARGUMENTS` to completion. 1. Identify the target (first token of `$ARGUMENTS`). Open the matching milestone command and its playbook at `docs/playbooks/<target>.md`. If the target is ambiguous or absent, ask which milestone — do not guess. 2. Read first: `CLAUDE.md`, `docs/decisions/`, then the playbook and the specs it cites. Load the relevant skills and say which. 3. Follow the `milestone-workflow` skill: restate goal + gate, check preconditions, work the ordered workstreams, capture forward-looking data per the project's capture plan, self-v...

Drive the milestone named in $ARGUMENTS to completion.

Identify the target (first token of $ARGUMENTS). Open the matching milestone command and its playbook at docs/playbooks/<target>.md. If the target is ambiguous or absent, ask which milestone — do not guess.
Read first: CLAUDE.md, docs/decisions/, then the playbook and the specs it cites. Load the relevant skills and say which.
Follow the milestone-workflow skill: restate goal + gate, check preconditions, work the ordered workstreams, capture forward-looking data per the project's capture plan, self-verify, update context in the same PR.
Spawn subagents for independent tracks where the playbook calls for them; run code-reviewer and doc-sync before declaring done.
Stop only at the playbook's documented decision points or a real ambiguity.

Treat the remaining text in $ARGUMENTS as scoping notes for this run.

Drive the milestone named in $ARGUMENTS to completion.

Identify the target (first token of $ARGUMENTS). Open the matching milestone command and its playbook at docs/playbooks/<target>.md. If the target is ambiguous or absent, ask which milestone — do not guess.
Read first: CLAUDE.md, docs/decisions/, then the playbook and the specs it cites. Load the relevant skills and say which.
Follow the milestone-workflow skill: restate goal + gate, check preconditions, work the ordered workstreams, capture forward-looking data per the project's capture plan, self-verify, update context in the same PR.
Spawn subagents for independent tracks where the playbook calls for them; run code-reviewer and doc-sync before declaring done.
Stop only at the playbook's documented decision points or a real ambiguity.

Treat the remaining text in $ARGUMENTS as scoping notes for this run.

Invocation

Configuration

Tool Access

Context Preview

Command Content