implement-spec

You have been provided a spec. This spec should have tickets associated with it, describing how to implement the spec.

The tickets are not a list of steps. They are a task graph with blocking relationships between them. This means there is always a frontier of tickets which are ready to be grabbed.

Communication to and from subagents should be sparse. Communicate primarily through context pointers: to the spec, tickets, and research notes. Don't duplicate information already available via pointers.

Implementer subagents should run in the background where possible, for maximum concurrency.

Steps

1. Read the spec and tickets, until you can name the current frontier and each ticket's blockers.

2. (optional) Use an exploration subagent to conduct any exploration required by the tickets - relevant codebase files or external documentation. If graphify is implemented (i.e. graphify-out/ exists), the exploration subagent should query it first to explore the codebase. Ensure the exploration subagent can save files - it should save its markdown notes in a directory outside the repo, accessible by all future subagents. This lets implementer subagents focus on implementation rather than exploration.

3. Use implementer subagents to implement each ticket on the frontier. Each implementer subagent should work in its own worktree, on its own branch.

4. As each implementer subagent completes, check whether it unblocks new tickets. Repeat step 3 for the updated frontier until no tickets remain.

5. (conditional) If any ticket created or changed UI, use an implementer subagent to run ./impeccable audit against it and implement the recommendations, then run ./impeccable audit again. Repeat until an audit round yields no more implementable recommendations.