Loading ToolBench
MCP server for extracting and processing subtitle files (.str format) and YouTube video transcripts
Both tools lack parameter type validation hints and error recovery guidance. No tool annotations (readOnlyHint, etc.) despite being read-only operations. Naming is acceptable (verb_noun pattern), but descriptions fail to guide LLM selection or explain when to use each tool.
Get all .str files in a given path.
Get subtitle of a given str file. Args: str_file_path: Full path to the str file. Returns: str: The subtitle of the str file. Raises: FileNotFoundError: Raises when the str file not found Exception: Raisers when running into parsing error of str file
Output schemas not documented. Neither tool declares what fields/structure it returns. LLMs cannot plan downstream operations or extract required data.
Lacks context for LLM selection and does not explain when to use it vs other discovery tools.
No tool annotations despite read-only semantics. Tools should declare readOnlyHint=true to signal safety to agents and enable optimizations.
Error handling lacks recovery guidance. get_subtitle raises FileNotFoundError and generic Exception but does not tell the LLM what to do next (e.g., 'Try get_all_subtitle_files() to discover valid paths').
Inferred effective spec: <=2025-11-25.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 38 | <=2025-11-25 | v2 |
Parameter 'path' in get_all_subtitle_files lacks validation hints. No description of expected format (absolute vs relative), constraints (must exist, must be readable), or error cases.