validate-findingslisted

# Validate Findings Opt-in coordinator-only step that runs an LLM validation pass over a consolidated v3 report. Adds AI assessment, verdict, and confidence to each finding. Under the v3 contract producers emit `risk`/`impact`/`scope` themselves, so the typical run leaves the floats untouched; this skill only re-estimates them in the rare case the consolidator left them absent (partial producer output that still satisfied the schema). NOT part of the automatic review pipeline — invoke after `consolidate_reports.py assemble` when a triage-quality validation pass is wanted. **Argument**: `$ARGUMENTS` — path to the consolidated `report.json` to validate. Edited in place. ## Inputs - A consolidated v3 report on disk (output of `consolidate_reports.py assemble`). - The producer commit, when `metadata.commit` is present, for best-effort source lookup via `git show`. ## Per-finding loop For each finding that does not already carry `ai_verdict`: 1. **Read context** — pull `description`, `recommendation`, any `code_snippets` (when absent, work from `description` alone), and optionally `git show <metadata.commit>:<path>` for the file referenced by `location`. Skip the `git show` lookup silently when `metadata.commit` is absent (non-git directory) or the command fails. 2. **Validate** — judge whether the finding holds against the code. Produce: - `ai_assessment` (Markdown) — rationale: what was checked, what was found, what the verdict turns on. - `ai_verdict` — one of `val