← ClaudeAtlas

titan-rtklisted

Evaluate rtk-ai/rtk for Claws integration. EVALUATION-ONLY titan — produces integration plan + stub OR rejected-eval card with reasoning. Opus LEAD, 4-6h ceiling.
neunaha/claws · ★ 9 · AI & Automation · score 76
Install: claude install-skill neunaha/claws
# Titan: rtk ## Mandate EVALUATE rtk-ai/rtk for integration into Claws. Decision titan — output is either: (a) **Adopt plan** — integration design + stub `lib/rtk/**` + skill family + dispatch manifest for a v0.8.4 follow-up wave, OR (b) **Reject card** — explicit rejected-eval at `.local/blueprints/v083-titan-rtk/REJECTED.md` with reasoning matrix (capability overlap, maintenance burden, license fit, cost-vs-value). Reference: https://github.com/rtk-ai/rtk — WebFetch in AUDIT phase. Per [[separate-goal-from-implementation]], enumerate ≥3 alternative paths before recommending one. ## Deliverables (conditional) **If adopt:** | Artifact | Path | |---|---| | Integration plan | `.local/blueprints/v083-titan-rtk/INTEGRATION-PLAN.md` | | Capability matrix | `.local/blueprints/v083-titan-rtk/capability-matrix.md` | | Stub module | `lib/rtk/stub.js` (NEW — interface-only) | | Skill | `.claude/skills/claws-rtk-eval/SKILL.md` (NEW) | | Staging changelog | `CHANGELOG_TITAN_RTK.md` | | READY-REPORT | `.local/blueprints/v083-titan-rtk/READY-REPORT.md` | **If reject:** | Artifact | Path | |---|---| | Rejected eval card | `.local/blueprints/v083-titan-rtk/REJECTED.md` (reasoning matrix) | | Capability matrix | `.local/blueprints/v083-titan-rtk/capability-matrix.md` | | Staging changelog | `CHANGELOG_TITAN_RTK.md` (NOTE-ONLY entry) | | READY-REPORT | `.local/blueprints/v083-titan-rtk/READY-REPORT.md` (with rejection score ≥80/100 = "high confidence rejection") | ## Bus topics | To