sttConvert a speech audio file to text (STT). 음성 파일을 텍스트로 변환합니다. MP3, WAV, M4A, AAC, OGG, FLAC, WEBM 등 일반적인 오디오 포맷을 지원하며, 변환된 텍스트를 JSON으로 반환합니다. [호출당 50포인트]
| Name | Required | Description | Default |
|---|---|---|---|
| language | No | 추출 언어 코드 (예: ko, en, ja). 기본값 ko | |
| audio_url | Yes | 다운로드 가능한 https URL (허용 형식: audio/mpeg, audio/mp3, audio/wav, audio/x-wav, audio/mp4, audio/aac, audio/ogg, audio/flac, audio/webm) (최대 200MB) |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations declare readOnlyHint=true, and the description adds useful behavioral details: supported audio formats (MP3, WAV, etc.), return format (JSON), and a 50-point cost per call. This goes beyond the basic read-only safety hint.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is short and front-loaded, but includes the same information in both English and Korean, creating slight redundancy. It remains efficient overall.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the simple nature of the tool and no output schema, the description mentions the return type (JSON) but not the exact structure or fields. It covers supported formats and cost, making it fairly complete but missing response shape details.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is 100%, with detailed descriptions for both audio_url and language. The tool description adds no new parameter semantics beyond what the schema already provides, so baseline 3 applies.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the tool converts speech audio to text, using a specific verb ('Convert') and resource ('speech audio file'). This distinguishes it from siblings like tts (text-to-speech) and voice_change.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
It provides clear context: use when you need to transcribe audio files. However, it does not explicitly mention when not to use it or compare with alternatives such as tts, so no exclusions are given.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Add one secure layer between your agents and this server.
Each tool has a clearly distinct purpose, including the watermark pair: draw_watermark_image is visible text while set_watermark embeds an invisible code. The TTS job lifecycle tools are also cleanly separated by action and output type.
The set mixes conventions: conversion tools use input_to_output, watermark tools use verb_noun, TTS jobs use a tts_jobs_ prefix, and stt is a bare acronym. The names are readable but do not follow one predictable pattern.
At 19 tools, the server sits in the borderline 16-25 range and spans document conversion, image processing, watermarking, audio/video, and async TTS. Most tools earn their place, but the overall surface feels somewhat heavy for a single conversion-focused server.
The server covers its core domains well: document conversions, watermarking with both visible and invisible methods, PDF operations, and a full async TTS workflow. Minor gaps exist, such as missing image-to-PDF or Excel-to-JSON inverse conversions, but agents can generally complete workflows without dead ends.