goal-workflowlisted

Run a settled implementation goal as a bounded autonomous build loop — lock the goal from the conversation and any docs written this session, front-load every decision, map the terrain, write a completion-invariant contract before any code, then loop (building with explicit Agent-tool fan-out, committing at intervals, verifying at milestones with fresh-eyes against the contract) until the invariants hold, and close out. Gated on a `--confirm` flag asserting the user has manually set ultracode effort and auto-accept mode (a skill can set neither); without the flag it stops, gives the setup steps, and offers two paths — a managed `--confirm` run, or a copy-pasteable `/goal` command that hands the work to native goal + workflow orchestration. An optional `--commit` flag (default off) enables commit-at-intervals and push-at-milestones in both the managed run and the generated command; without it, nothing is committed. Use when a plan is settled and you want Claude to implement it end-to-end on its own — typically
adamlinscott/claude-skills · ★ 3 · AI & Automation · score 76

Install: claude install-skill adamlinscott/claude-skills

# Goal Workflow Turn a settled goal into a **bounded implementation loop, then force proof, then fix or close.** This is the implementation-phase sibling to `assumption-inventory` (before), `reground` (mid-drift), and `fresh-eyes` (after). The orchestration engine is **explicit subagent fan-out via the Agent tool** — the only multi-agent mechanism a skill can reliably drive. Native `/goal` and ultracode's automatic workflow orchestration are **user-only and discretionary** (a skill cannot set `/goal`, and ultracode decides for itself whether to author a workflow), so this skill does not depend on them — it offers them as an optional fast path the user can run. `/fresh-eyes` is the independent verifier. The skill's own job is to lock the goal, front-load decisions, map terrain, and — the load-bearing step — **write a checkable contract before any code, so verification anchors to a bar set in writing, not to the implementer's memory of intent.** Run the steps in order. Step 0 is a hard gate; do not skip it. ## 0. Confirmation gate (FIRST — before anything else) This skill runs a long, expensive, autonomous loop, and it depends on two session settings a skill **cannot** set or detect for itself: **ultracode effort** (xhigh + workflow orchestration) and **auto-accept mode** (so the loop's agents run without a permission prompt on every action). Passing the keyword `ultracode` as an argument does **not** turn effort on — only the user can, manually. So gate on an explicit `--