awesome-writing-plans

Awesome Writing Plans

Overview

Project-level wrapper for superpowers:writing-plans that adds a fresh-subagent rigor review loop after the plan is written and before handoff to executing-plans. Each round asks "what realistic failure mode would survive these tests?" and the main agent either extends the plan with a closing test or marks the gap out-of-scope with a written rationale. Loop runs up to 3 rounds; terminates on 0 critical gaps or escalates to the user with a summary.

Announce at start: "I'm using the awesome-writing-plans skill — write the plan with a ≤3-round test-rigor pass before handoff."

Core principle: Run upstream writing-plans → Pass B rigor loop → extend or defer each critical gap → handoff to executing-plans.

The Process

digraph awesome_writing_plans {
    rankdir=TB;
    upstream [label="Step 1: Run\nsuperpowers:writing-plans" shape=box];
    written [label="Plan written" shape=box];
    self_review [label="Upstream Self-Review\n(inline checklist)" shape=box];
    round [label="Step 2: Pass B Round N\nDispatch subagent" shape=box];
    gaps [label="Critical gaps\nreturned?" shape=diamond];
    address [label="Extend plan OR\nadd to Deferred Risks" shape=box];
    cap [label="Round count\n≥ 3?" shape=diamond];
    escalate [label="Escalate to user\nwith summary" shape=box];
    handoff [label="Step 3: Handoff to\nexecuting-plans" shape=doublecircle];

    upstream -> written;
    written -> self_review;
    self_review -> round;
    round -> gaps;
    gaps -> address [label="yes"];
    gaps -> handoff [label="no"];
    address -> cap;
    cap -> escalate [label="yes"];
    cap -> round [label="no"];
    escalate -> handoff;
}

Step 1: Run upstream writing-plans

Invoke superpowers:writing-plans.

Run it to completion through its own "Self-Review" step. Do not proceed to the upstream skill's "Execution Handoff" step until Pass B runs (Step 2). The override inserts between self-review and execution-handoff.