Instructions file GitHub Copilot
Copilot instructions for The-AI-Alliance/tapestry, covering copilot instructions for project tapestry, about the project, general review guidance, reviewing contrib/ contributions and scope & neutrality.
Instructions file CodexOpenCode
AGENTS.md instructions for The-AI-Alliance/tapestry, covering tapestry repo guide, repository map, where to look first, docs map and python package layout.
Command Claude Code
Audit openspec/specs/ against the current source code and update specs that have drifted.
Skill Claude CodeCodex
First-pass triage for a proposed change to CUBE's shared contract (an RFC / OpenSpec change / API request). Separates the contributor's real need from the mechanism they proposed, finds the smallest thing that serves it, and produces an advisory verdict (RESHAPE / DECLINE / ESCALATE / ACCEPT) plus a contributor-facing…
Skill Claude CodeCodex
Interview the user and scaffold a new CUBE benchmark end-to-end — from requirements discovery through cube init, filling template TODOs, pytest + debug-suite validation, and submitting to the cube registry. Invoke when the user runs /new-cube or asks for help creating a new CUBE benchmark.
Skill Claude CodeCodex
Review a CUBE benchmark submission — validate the registry YAML (if present), audit the cube code against cube-standard invariants, install the package, run pytest tests/ and cube test , and produce a structured Blocking / Suggestions report. Accepts a local path, a cube-registry-entry.yaml path, or any GitHub PR URL…
Instructions file CodexOpenCode
AGENTS.md instructions for The-AI-Alliance/cube-standard, a project described as: Standardize benchmark wrapping so the community can wrap various otherwise-incompatible benchmarks uniformly and use them everywhere.
Instructions file
Claude Code instructions for The-AI-Alliance/cube-standard, covering claude.md — cube-standard, what this repo is, routing contributors (point them to the right place), the 5-layer architecture and engineering principles.