feat(skills): add OpenTUI performance workflow

[benvinegar]

Summary

• bundle a new opentui-performance agent skill for profiling and optimizing React/OpenTUI terminal applications
• cover rendering and host-object cost, deterministic geometry and hierarchical windowing, workers/caches/publication scheduling, memory, benchmark integrity, PTY/TTY validation, and terminal text correctness through progressive reference documents
• include a measured Hunk case study while keeping Hunk-specific policies and thresholds out of the generic workflow
• expose the skill through hunk skill path opentui-performance and the shorter hunk skill path performance alias
• package the complete skill and references in npm and standalone artifacts, while tightening prebuilt npm staging to exclude maintainer-only skills

Skill structure

skills/opentui-performance/
├── SKILL.md
└── references/
    ├── rendering-and-geometry.md
    ├── async-and-memory.md
    ├── benchmarking-and-validation.md
    └── hunk-case-study.md

Validation

• skill frontmatter and all progressive-disclosure links validated
• independent skill and packaging review — no remaining findings
• bun run typecheck
• bun run lint
• bun run format:check
• bun run test — 3,013 passed, 10 skipped
• bun run check:pack
• bun run build:npm
• host prebuilt staging and bun run check:prebuilt-pack
• rebuilt host binary and bun run smoke:prebuilt-install
• bun run changeset:status

Draft questions

• Is bundling this with Hunk the right long-term distribution, or should the generic skill eventually live in its own skill repository?
• Is the Hunk case study the right amount of product-specific evidence, or should it be shortened further?
• Should a future revision include executable benchmark scaffolding, or remain methodology-first and adapt to each repository's harness?

This PR description was generated by Pi using gpt-5.6-sol

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

1 Skipped Deployment

Project	Deployment	Actions	Updated (UTC)
hunk-web	Ignored	Preview	Aug 19, 2026 2:01am

[greptile-apps]

Greptile Summary

This PR adds and distributes an OpenTUI performance-engineering skill, exposes it through named CLI lookups, and narrows prebuilt npm staging to public bundled skills.

• Adds the performance workflow and four progressive reference documents.
• Adds opentui-performance and performance CLI lookup names with colocated tests.
• Includes the complete skill in source npm, prebuilt npm, and standalone release artifacts.
• Updates packaging checks and installation smoke coverage while excluding maintainer-only skills from prebuilt npm staging.

Confidence Score: 5/5

The PR appears safe to merge; no concrete changed-code defect remains.

The CLI catalog, source package, staged prebuilt package, standalone artifacts, and their validation paths consistently include the new skill and references, with tests covering canonical and alias lookup behavior.

Important Files Changed

Filename	Overview
src/core/run/paths.ts	Adds the new bundled skill and short alias to the shared catalog used by runtime lookup and packaging.
src/app/cli.ts	Exposes the performance skill in CLI help and routes its canonical name and alias through existing path resolution.
scripts/stage-prebuilt-npm.ts	Replaces broad skill-directory copying with recursive staging of the public bundled-skill allowlist.
scripts/check-prebuilt-pack.ts	Verifies the performance skill and references are packaged and rejects maintainer-only skill files.
scripts/smoke-prebuilt-install.ts	Exercises every canonical bundled skill name, both relevant aliases, and an installed performance reference.
skills/opentui-performance/SKILL.md	Defines the progressive OpenTUI profiling and optimization workflow with links to the bundled references.
.github/workflows/release-prebuilt-npm.yml	Extends standalone release validation to require the new skill and all four reference documents.

_{Reviews (1): Last reviewed commit: "feat(skills): add OpenTUI performance wo..." | Re-trigger Greptile}