spec-driven-sprintlisted

# Spec-Driven Sprint Turn one GitHub issue into a self-contained, **resumable** spec-driven sprint: research the fix surface, materialize an HTML spec (work cards + embedded AI resumption prompt), branch an isolated worktree off the remote default, execute each card with TDD where there is code, QA-validate against the spec's acceptance criteria in an iteration loop, then open a PR. No story file is produced — this skill is **ad-hoc and issue-driven**, mirroring `spec-driven-dev`'s Execute-Verify-Gate discipline without `spec-driven-dev`'s story-scoped artifacts. **If ambiguous or conflicts detected: HALT and use AskUserQuestion.** --- ## Portability — consumer-safe core + DevForgeAI addendum This skill runs in **any** repository, not only DevForgeAI's own. Its phase files keep a **portable core** (generic / runtime-detected: default branch, required check, test runner) and apply a **DevForgeAI addendum** only when both sentinels are present: - `devforgeai/specs/context/` (the constitutional context folder), and - `.github/workflows/python-cli-tests.yml` (the required-check workflow). When the sentinels are absent, the host-specific addendum mechanics are skipped in favor of the runtime-detected equivalents (default branch, required check, test runner). The split rule-set lives in `references/portable-workflow.md` (consumer-safe core) and `references/devforgeai-addendum.md` (sentinel-gated host mechanics — worktree placement, the required-check name, coverage/drift ga