AI不止语
Community@jnmetacode
AI 编程实战 · AI Coding Practitioner · 公众号 AI不止语
Agent Skills by AI不止语
Showing 20 vetted skills indexed across 1 GitHub repositories.
using-git-worktrees
Create isolated Git worktrees with ignore validation and setup steps.
test-driven-development
Enforce a red-green-refactor workflow with failing tests before production code.
systematic-debugging
Identify root causes of bugs through a four-phase debugging workflow.
chinese-git-workflow
Organize domestic Git workflows for Gitee, Coding.net, and Kirin GitLab.
using-superpowers
Invoke the Skill tool at interaction start to determine applicable skills.
dispatching-parallel-agents
Dispatch parallel autonomous agents to diagnose and fix independent failure domains.
chinese-code-review
Convert imperative code review comments into constructive, culturally aware suggestions.
executing-plans
Execute written plans with review checkpoints and auditable task progress.
finishing-a-development-branch
Guide Git branch finalization through merge, PR, or cleanup options.
chinese-documentation
Enforce Chinese technical documentation standards for typography, terminology, and formatting.
brainstorming
Guide exploratory brainstorming sessions to convert ideas into actionable design briefs.
workflow-runner
Execute agency-orchestrator YAML workflows using the current session's LLM.
writing-plans
Generate executable implementation roadmaps for multi-step software features.
requesting-code-review
Automate peer review of code changes using git range references and code-reviewer templates.
receiving-code-review
Translate ambiguous code-review feedback into precise, verifiable action plans.
writing-skills
Create and maintain Skills documentation with test-driven development workflows.
chinese-commit-conventions
Standardize Chinese commit messages with conventional commit formatting and issue linkage.
mcp-builder
Generate TypeScript and Python MCP server projects with structured tooling and validation.
verification-before-completion
Require verification command evidence before declaring task completion.
subagent-driven-development
Delegate independent tasks to fresh subagents with two-stage reviews.