claude-better-plan

Executes a plan produced by /deep-plan: the plan-folder artifact whose plan.md carries a ## Decisions made table and per-task **Tests (TDD)** blocks, located when no path is given through the approved-plan memo recorded at approval. Parses the plan's ## Tasks into harness tasks (one TaskCreate each), wires Depends on into addBlockedBy, then drives a test-first implementation loop task by task in dependency order. Invoke after a /deep-plan plan is approved and you are ready to build it, e.g. "implement the plan" or "/deep-plan:deep-plan-execute <plan-file>". Not for executing plans produced outside /deep-plan.

Triangulated, decision-surfacing planning for non-trivial tasks. Fans research three ways in parallel, surfaces every meaningful sub-decision as an AskUserQuestion with 3 to 5 options, runs targeted deep web research per chosen option, and produces an AI-consumable plan file with TDD-embedded tasks. Triggers on the slash command /deep-plan only.

Standalone design review of code, diffs, or plan files by a parallel critic fleet: one small-model critic per red-flag cluster (module depth, information hiding, naming, comments), then adversarial verification. Use when the user asks to "design-review" something, review design quality, check for shallow modules or information leakage, or audit a diff or plan against design principles.

Standalone review of the Tests (TDD) blocks in a plan produced by /deep-plan, run by a parallel critic fleet: one small-model test critic per red-flag cluster (assertions and failure output, test doubles and seams, fixtures and setup, test level and duplication), then adversarial verification. Use when the user asks to review planned tests, check TDD quality, or assess assertions, doubles, fixtures, or flaky-test craft in a deep-plan plan file. Not for reviewing diffs, pull requests, or already-implemented code.