/ce-work

/ce-work — “Execute against implementation-ready guardrails, figure out HOW with code, ship through quality gates.” Step 3 of the loop and the “20%” of the effort split. It takes the guardrailed plan from ce-plan and implements it in code — the plan owns the WHAT, ce-work owns the HOW — running inside an isolated git worktree (pattern-worktree-isolation, via ce-worktree) and using MCP tooling (e.g. Playwright, XcodeBuildMCP) to simulate real usage while building.

It implements stage-implement and is the seventh member of the wiki’s execute cluster.

Cross-framework equivalents

Execute cluster (now seven frameworks): ce-work ↔ gsd-execute-phase ↔ openspec-apply ↔ addy-incremental-implementation ↔ speckit-implement ↔ bmad-dev-story — turn a plan into working code. ce-work’s distinctive constraints: guardrails from the plan bound it, and it runs in worktree isolation with live-usage simulation.

See Also

• gsd-execute-phase · openspec-apply · addy-incremental-implementation · speckit-implement · bmad-dev-story — execute-cluster counterparts.
• ce-worktree — the isolation it runs inside; ce-plan — supplies its guardrails.
• ce-simplify-code · ce-code-review — the next loop steps that gate its output.
• stage-implement — the canonical stage this implements.