개발 스킬

Use `uv` to run Python code, sync project dependencies, manage lock files, select Python versions, and run tools. The skill helps choose commands for each task and clarifies whether they change the project environment or lock file.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Running a Python script with temporary dependencies
• Syncing or checking a project environment
• Checking or updating a lock file

Create and debug ast-grep rules to find code by its syntax structure rather than plain text. Use patterns and relational or combined conditions to locate precise code shapes, then support targeted searches and refactors across a project.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Finding a specific syntax pattern across a codebase
• Searching for code nested inside or containing another construct
• Combining conditions to narrow structural searches

Build and troubleshoot Tauri v2 desktop apps, from project setup and connecting a web frontend to Rust commands through permission configuration, plugin integration, and cross-platform packaging. Get ordered commands and focused checks for the task at hand.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Starting a Tauri v2 project from scratch
• Connecting a web frontend to Rust commands with invoke
• Setting limited fs, http, or shell permissions

Diagnose and recover Trae’s native custom-model route when requests fail, error codes appear, or an update changes the DLL hook offsets. It focuses on the existing URL rewriter and helps avoid repeating approaches that previously failed.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Investigating Trae custom-model errors such as 4028, 4054, or 429
• Recovering the native route after a Trae update changes hook offsets
• Diagnosing URL rewriting in the native rewriter or SseOpenPayload path

Write and troubleshoot PowerShell commands that work across Windows PowerShell 5.1 and PowerShell 7. It helps handle quoting, pipes, native command arguments, and version differences, with commands and compatibility notes for the target version.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Passing a jq filter with a pipe to an external command
• Fixing commands that parse arguments incorrectly
• Adapting scripts for differences between PowerShell versions

Automate browser tasks from the command line, including navigating pages, finding controls, filling forms, and collecting page information. Compact snapshots help agents work through long sessions, while screenshots and video recordings support debugging and cloud browser testing.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Automating a multi-step website workflow
• Filling and submitting web forms
• Capturing screenshots or recordings for debugging

Set up an agent work harness for a new or existing project. It detects the stack, reviews current setup, scaffolds process directories, and deep-scans the codebase to write project context and test guidance, asking before reorganizing existing files.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Preparing agent workflows for a new project
• Refreshing setup for a project with existing context
• Mapping architecture, tools, and test commands

Structure a two-layer review of a plan: four cross-cutting checks and a feasibility review for each section. The findings are synthesized into a PASS, CONDITIONAL, or BLOCKED gate verdict, with gaps and proposed plan updates made explicit.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Reviewing a plan before execution
• Checking infrastructure, tests, changes, and security
• Finding missing steps or conflicts in plan sections

Create a read-only snapshot of repository activity, including branches, worktrees, and active plans. The skill organizes recent work, likely plan context, warnings, and suggested checks into a handoff summary without changing project state.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Getting oriented at the start of a session
• Preparing a handoff for another agent
• Reviewing active plans and recent work

Investigate bugs, failing tests, CI problems, and slow systems by tracing evidence to the root cause before proposing a fix. It supports code and system analysis, severity-aware verification, and clear diagnostic reports.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Finding why a test or build fails
• Tracing an error through a call stack
• Investigating CI or server incidents

Turn an idea, product requirements document, or approved direction into a saved implementation plan. The skill selects a SIMPLE or COMPLEX format, includes scope and verification details, and structures larger programs as an umbrella plan with separate phase plans.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Turning a feature idea into a saved implementation plan
• Writing a plan from an approved product requirements document
• Creating an umbrella plan for a multi-phase program

Prepare kickoff materials for a program with three or more dependent phases. It structures an umbrella plan, a goal charter, session-goal block, phase plan stubs, and a repeatable seven-step workflow with validation checkpoints.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Planning a project that spans several dependent phases
• Defining validation gates before each phase begins
• Starting work across multiple packages or services

Look up library and framework documentation for API, configuration, setup, or version questions. It queries Context7 first, then uses llms.txt discovery scripts when coverage is missing or repository-oriented research is requested.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Find the correct API usage for a library
• Check setup or configuration instructions
• Look up version-specific documentation

Review a proposed feature or risky change before implementation through five expert perspectives: architecture, security, performance, UX, and devil’s advocate. The skill compares concerns, identifies risks, and returns a GO, CAUTION, or STOP verdict with recommendations.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Checking a major feature before implementation
• Comparing competing technical approaches
• Finding risks in a significant refactor

Quickly locate files and understand where functionality lives in a codebase. It starts with shell searches and can optionally split a large search across research agents or external Gemini/OpenCode tools, then returns a concise report of relevant files and open questions.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Finding files related to a feature across directories
• Locating where a function or behavior is implemented
• Gathering codebase context before debugging

Check whether Claude and Codex agent definitions, skill registries, README.md, and development protocol files stay in sync. The validators identify wiring gaps and inconsistencies, then guide small fixes and rechecks.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Checking Claude and Codex agent definitions after a change
• Finding skill registry or invocation wiring gaps
• Verifying README and protocol files after a move

Create or refresh the project’s shared context documents from the current repository state. It records architecture, tooling, plans, and relevant sources in all-context.md and grouped context files, then runs validation scripts to check the result.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Creating context when project documentation is missing
• Refreshing context after code or tooling changes
• Updating a specific package or feature area first

Finds files relevant to a task by scanning project context, protocols, plans, and feature folders. It routes by task domain or matching keywords, surfaces full paths and frontmatter details, and provides a consistent starting point for an agent session.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Finding project guidance for a task
• Locating plans and feature files by name
• Routing to context documents using task keywords

Find plan artifacts related to the current task and group them as active, backlog, completed, reports, or references. It checks the current feature in depth, other features' active plans, and general active plans to reveal prior work and deferred items.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Finding plans related to a feature
• Checking what work was tried before
• Reviewing deferred tasks and completed plans

Create a post-execution closeout for a plan or phase. It classifies archive readiness, scores signals that call for process updates, recommends a commit checkpoint, records what was verified, and names the single best next state or follow-up plan.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Closing out a completed implementation plan
• Recording phase evidence before moving to the next phase
• Deciding whether a plan is ready to archive

Check whether project context documents and shared skills are discoverable and routed correctly across Claude and Codex. The skill runs validators for missing links, metadata, dependencies, and catalog drift, then helps identify the smallest relevant fixes.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Checking context routing after documentation moves
• Finding shared skills missing from discovery paths
• Verifying Claude and Codex agent wiring

Guides contributors through Tel-Agent's contribution process, from choosing an available issue and setting up the repository to making a focused change, checking it, and opening a pull request. It also explains the project's translation workflow for first-time contributors.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Finding an issue that is ready to work on
• Setting up a fresh Tel-Agent checkout
• Preparing a change for review

Explains how to add or change a messaging channel in Tel-Agent, including its transport, manifest, setup descriptor, settings card, tests, health reporting, and documentation. It covers credential handling, webhook verification, message delivery, and channel-specific checks.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Adding a new messaging channel
• Changing an existing channel transport
• Implementing webhook signature checks

Turn a vague idea into a clear goal, constraints, and a list of unknowns before specification work begins. The skill asks one question at a time, recommends an answer for the user to consider, and hands the structured results to the spec skill.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Clarifying an early idea before writing a specification
• Stress testing a plan through focused questions
• Separating confirmed constraints from unknowns