tritonkit-ops-governance

Governs TritonKit development loops, documentation memory, and session consolidation workflows.

1|Updated May 15, 2026
One-click install
npx skills add https://github.com/NeptuneKit/TritonKit --skill tritonkit-ops-governance-neptunekit
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: tritonkit-ops-governance
Source: https://github.com/NeptuneKit/TritonKit/tree/main/.agents/skills/tritonkit-ops-governance
Command: npx skills add https://github.com/NeptuneKit/TritonKit --skill tritonkit-ops-governance-neptunekit

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Maintaining consistent development discipline across the TritonKit repository is hard: CLI/HTTP/Wails changes need BDD-first spaces, releases need strict gate ordering, and completed work sessions need structured consolidation into docs, memory, and skills. This Skill encodes those operational rules so an AI agent follows them automatically. ## Core Features & Use Cases - Development Loop Governance: Enforces BDD-first space creation under docs-linhay/spaces, Triton-first emulator actions, machine-readable CLI contracts, and fail-closed host-side fallbacks. - Release & CI Gates: Defines verify.sh local/CI entry points, Homebrew tap publishing, dual-architecture CLI packaging, and public skill tarball contracts. - Session Consolidation: When the user says "整理" or "沉淀", it isolates changes with git status, extracts reusable patterns into skills or AGENTS.md, updates docs-linhay memory, and creates a scoped commit. - Use Case: After finishing a multi-issue worktree session, ask the agent to consolidate; it will check each worktree, merge cleanly, run gates, update INDEX.md and memory, and close verified GitHub issues. ## Quick Start Ask the agent to consolidate and commit the just-finished TritonKit work session following the ops governance rules.

Frequently Asked Questions about tritonkit-ops-governance

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I consolidate a finished TritonKit work session?▼

Say "整理" or "沉淀" and the agent runs git status and git diff --stat to isolate changes, extracts reusable patterns into skills or docs-linhay memory, runs check-docs.sh, and creates a scoped commit without git add -A.

What is the Triton-first rule for emulator automation?▼

Before calling baguette, xcrun, simctl, hdc, adb, or raw xcodebuild, the agent must save machine-readable output from triton status, doctor, capabilities, schema, or plan. Fallback is allowed only when Triton returns failure or unsupported, with evidence preserved.

How are TritonKit releases published to Homebrew?▼

Maintainers run docs-linhay/scripts/release.sh with a version tag. CI builds arm64 and cross-compiled x86_64 CLI tarballs, checksums, and the tritonkit-skills.tar.gz bundle, then pushes the formula to NeptuneKit/homebrew-tap.

When should rules go into AGENTS.md versus a skill?▼

Only repo-wide, long-term stable rules that must always be followed belong in AGENTS.md. Domain- or process-specific reusable actions go into an existing skill first; new skills require deciding whether they are public or internal.

Why avoid git stash when running parallel worktrees?▼

Git stash is repository-global and shared across worktrees, so stash push or pop in one worktree can pull in another worktree's WIP or orphan references. Use temporary branches or separate files for intermediate state instead.