'json output' 검색 결과

Build workflows that extract structured data from LLM responses, check field types and values with Pydantic, and retry failed extractions. Includes examples for nested data, classification, and receiving results as they are generated.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Extracting contact and company details from text
• Classifying articles into predefined categories
• Checking review analysis against validation rules

Build a new agent skill from a structured template. Copy the directory and edit SKILL.md to document its purpose, inputs, outputs, processing steps, precautions, and examples, replacing placeholders with concrete content.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Draft the documentation for a new agent skill
• Define a skill's inputs and output format
• Organize processing steps and precautions

Check whether a file creation or transformation task actually succeeded after an unclear execution error. Inspect output files and their contents to distinguish success, partial completion, and failure before deciding whether to retry.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Checking output files after an unknown error
• Validating spreadsheet or JSON contents
• Deciding whether to retry or fix incomplete output

Run Python through run_shell when code execution tools repeatedly fail. This workflow embeds a multiline script directly in a shell command, supporting tasks such as CSV conversion and PDF generation while checking execution output.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Continuing Python work after repeated tool errors
• Converting CSV data to JSON
• Generating PDFs with reportlab

Use shell heredocs, which pass multiline text directly to a command, when write_file fails on complex code. Save TypeScript, JSX, JSON, or shell scripts with literal quotes and formatting, then verify the file and its contents.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Saving TypeScript or JSX files after repeated write_file failures
• Writing JSON files containing quotes and special characters
• Preserving variables and line breaks in shell scripts

When Python execution repeatedly fails in execute_code_sandbox, use a file-based alternative: save the script, run it through the shell, and capture results or errors. The guidance covers paths, imports, timeouts, structured output, and optional cleanup.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Retry Python work through the shell after repeated sandbox failures
• Run longer tasks with an adjusted timeout
• Execute Python work that needs multiple files or system utilities

Retrieve concise summaries from GWAS Catalog, a resource for studies linking genetic variants to traits. Look up studies, associations, genes, variants, and publications, with JSON output or saved full responses when needed.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Finding GWAS studies related to asthma
• Checking associations mapped to BRCA1
• Looking up genetic variants and related publications

Build n8n workflows that download, transform, and send files without losing their contents. Learn how file data differs from JSON, how to preserve attachments, and how to use storage keys or URLs with AI tools and chat messages.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Fixing attachments lost during data transformation
• Passing uploaded files to AI agent tools
• Displaying generated images in chat

Investigate unexpected Claude Code output by matching a session’s local transcript with production router logs. Identify the serving model and trace response conversion code to explain empty blocks, tool output, or other rendering symptoms.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Investigate unexpected output in a specific session
• Examine an empty thinking block
• Match displayed content to the serving model and provider

Check an academic paper’s tables, numerical claims, and modeling choices against R scripts and outputs. Produce discrepancy reports, a record linking claims to code, and automated tests that rerun the analysis.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Checking table values against R outputs
• Tracing numerical claims to supporting results
• Reviewing sample sizes and model specifications

Draft huggingface_hub release notes from saved pull request JSON files. The skill groups changes, explains major updates, adds examples from documentation changes, and includes author credits and verified documentation links in a Markdown document.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Drafting release notes from saved pull request data
• Explaining major features and breaking changes
• Adding command or API examples from documentation changes

Find lemmas, reusable mathematical results, in Mathlib by matching the type signature you need for a Lean proof. Search with type variables, related terms, or exact names to identify candidates you can apply.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Finding a lemma when its name is unknown
• Exploring lemmas associated with a particular type
• Finding list functions from their input and output types

Provides a reference for building Claude Code hooks, scripts or model checks triggered by events. Covers input and output formats, registration, testing, and debugging for tasks such as blocking tool calls, adding context, and checking tests before stopping.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Create a hook that blocks edits to protected files
• Format files automatically after edits
• Add Git context when submitting a prompt

Expert guide for agent-native structured outputs, JSON Schema strict mode enforcement (OpenAI response_format, Gemini schema mode, Claude tool_use), MCP resource schemas, and zero-hallucination type contracts / Panduan ahli kontrak output terstruktur native agen dan penegakan JSON schema ketat.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

Hook JSON input/output API reference — what data hooks receive via stdin and what JSON they can return to control Claude Code behavior. Use when writing hook scripts, checking exit code behavior, building JSON output for PreToolUse permissions, or understanding event-specific input schemas.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

Add --json flag to CLI commands for machine-readable output on stdout. Generic pattern for any CLI tool, enabling piping to jq and CI/CD integration while keeping human-readable output on stderr.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

Perl 코드의 입력 처리, 파일 작업, 외부 명령, SQL과 웹 출력을 점검합니다. 구체적인 예시와 점검표에 따라 안전한 구현으로 수정합니다.

• Perl 입력 검증을 점검합니다
• 파일 작업과 명령 실행을 확인합니다
• DBI의 SQL 삽입 공격에 대비합니다

대규모 언어 모델의 답변을 채점하는 평가 세트를 만들고 실행합니다. 진행 상황과 개별 답변의 통과 여부를 확인하고, 모델 비교나 CI의 품질 판정에도 활용합니다.

• 입력과 예상 답변으로 평가를 만듭니다
• 평가 진행 상황을 확인합니다
• 실패한 답변과 점수를 조사합니다

Manage work on a SeaClip-Lite project board through commands that create and update issues, control pipelines, configure schedules, and inspect recent activity, with JSON output for AI agents.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Find issues by status or priority
• Create issues and move them between board columns
• Start, resume, or stop work pipelines

Manage Obsidian notes through its Local REST API. Search a vault, create or update notes, append content, and execute Obsidian commands from the command line, with JSON output for agent workflows.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Find meeting notes in a vault
• Create project notes and append updates
• Find notes by metadata conditions

Manage Mailchimp audiences, create and schedule email campaigns, and review delivery results through commands. Agents can update subscriber records, inspect open and click rates, manage automated emails, and retrieve structured JSON data for further processing.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Organize subscribers with tags and segments
• Create and schedule email campaigns
• Review opens, clicks, and unsubscribes

Manage Node.js processes through PM2 from the command line. List running processes, start or restart applications, inspect logs and metrics, and save the process list, with structured JSON output for agent workflows.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Inspect running processes and their metrics
• Start, stop, or restart an application
• Review recent application logs

Delegate code analysis and implementation to Codex CLI without an interactive session. The skill reads ordered output events, extracts the final agent response, and reports errors, with a read-only sandbox used by default.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Review code in a read-only sandbox
• Delegate implementation to Codex
• Get a second coding agent's assessment

Delegate repository analysis and coding tasks to GitHub Copilot CLI without an interactive session. The skill reads structured output, checks errors, and summarizes the final result along with any denied tool requests.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Ask Copilot to explain a failing test
• Delegate repository analysis to Copilot
• Request code changes with narrowly scoped tool permissions