validate-issue-loop

# validate-issue-loop Chain validate-issue → (conditional) update issue → work-on-issue-loop into one autonomous run, so an issue goes from "reported" to "PR through N rounds of review" without a human in the loop between steps. This is validate-issue's normal interactive handoff (`→ Reply "work on issue"`) made unattended: the loop reads its own verdict and decides what to do next, instead of waiting for the user to type a reply. **Do not skip validation.** Auto-implementing an issue whose factual claims or proposal you haven't traced against the code just reproduces the issue's own mistakes in a PR. Every step of validate-issue still runs; only the "wait for the user's reply" step is replaced by a decision table. ## Input Same defaults as validate-issue: issue URL, `#<N>` / `<N>` / `owner/repo#N`, or nothing (defaults to the latest open issue in the current repo). ## Steps ### 1. Run validate-issue Invoke the `validate-issue` skill for the target issue (Skill tool, `skill: validate-issue`). Let it run its full process — steps 0 through 8 — and produce its verdict block: ``` **#<N>: Update issue description? <Yes|No>** · Complexity: <score>/100 — Capability <k> (<driver>); Volume <v> · fableplan: <yes|no> · Scope: <OK | too large — split/umbrella/narrow> ``` Its final `→ Reply "work on issue"...` line is written for interactive use — in this loop, treat the verdict block as structured output to parse yourself, not a prompt to wait on. Don't ask the user to confi...