gyc-12
Community@gyc-12 · @China
gyc-12 provides Trellis-based spec-driven development skills covering session bootstrapping, TDD, debugging, code review, and parallel task execution for coding agents.
Agent Skills by gyc-12
Showing 24 vetted skills indexed across 1 GitHub repositories.
trellis-spec-bootstarp
Generate project-specific Trellis coding specs from real codebase analysis.
trellis-before-dev
Loads project coding guidelines from .trellis spec indexes before implementation begins.
trellis-brainstorm
Guides collaborative requirements discovery and converges on MVP scope before implementation.
trellis-break-loop
Analyzes fixed bugs across five dimensions to produce prevention mechanisms and spec updates.
trellis-check
Verifies code changes against specs, lint, tests, and cross-layer consistency checklists.
trellis-meta
Guides customization of local Trellis workflow, hooks, agents, and platform files.
trellis-update-spec
Captures executable contracts and coding conventions into .trellis/spec/ documentation files.
trellis-start
Initializes Trellis development sessions by loading context, git status, and task routing.
trellis-finish-work
Archives completed tasks and records session journals after verifying committed work.
trellis-continue
Resumes in-progress Trellis tasks by routing to the correct workflow phase and step.
using-git-worktrees
Creates isolated git worktree workspaces with detection, setup, and baseline test verification.
test-driven-development
Enforces test-first development using the red-green-refactor cycle for features and bugfixes.
systematic-debugging
Diagnose bugs through a four-phase root cause investigation process before proposing fixes.
using-superpowers
Enforces skill discovery and invocation before any AI response or action.
dispatching-parallel-agents
Dispatch concurrent agents to fix independent test failures across separate problem domains.
executing-plans
Executes written implementation plans task-by-task with verification checkpoints and review gates.
finishing-a-development-branch
Guides completion of development branches through structured merge, pull request, or cleanup options.
brainstorming
Guides collaborative design dialogue to turn ideas into approved specs before implementation.
writing-plans
Generates detailed multi-task implementation plans with TDD steps from a spec.
requesting-code-review
Dispatches a code reviewer subagent to evaluate git diffs against requirements before merging.
receiving-code-review
Guides verification and implementation of code review feedback with technical rigor.
writing-skills
Create and test agent skills using a TDD-based RED-GREEN-REFACTOR documentation workflow.
verification-before-completion
Enforces running verification commands before claiming work is complete or passing.
subagent-driven-development
Executes implementation plans by dispatching fresh subagents per task with two-stage review.
Frequently Asked Questions About gyc-12
FAQPage SchemaWhat tasks can I accomplish with gyc-12's Trellis skills?▼
You can bootstrap project coding specs into .trellis/spec, brainstorm requirements into a PRD, execute plans with test-driven development, run systematic root-cause debugging, verify quality gates (lint, type-check, tests, spec compliance), and finish sessions with archival and journal logging.
Who should use these Trellis development skills?▼
Software engineers and coding-agent operators who want spec-driven, verifiable development sessions. They suit developers managing multi-phase tasks, parallel independent workstreams via subagents or worktrees, and teams enforcing code review and verification-before-completion discipline.
How does a typical session run with these skills?▼
Start with trellis-start to load git status, active tasks, and project guidelines, which routes to brainstorm, direct edit, or task execution. During work, trellis-continue resumes phases; at the end, trellis-finish-work verifies the quality gate, archives tasks, and records progress.
What prerequisites or dependencies do these skills require?▼
A project initialized with trellis init providing the .trellis directory, platform hooks, and settings. Spec bootstrapping optionally uses GitNexus, ABCoder, or direct source inspection, and trellis-continue relies on get_context.py with --mode phase for step-level detail.
How do these skills prevent repeated bugs and unverified claims?▼
trellis-break-loop analyzes root cause categories and why prior fixes failed, capturing prevention knowledge into specs. verification-before-completion requires running verification commands and confirming actual output before claiming work is complete, fixed, or passing.