Loading ToolBench

Reading the current corpus…

telegram-search MCP server · F · ToolBench

telegram-search

A Telegram message search and indexing server with AI-powered capabilities for searching, retrieving context, and analyzing chat data

2026-07-28+STDIO0 tools

45/100

Definition Quality0

Protocol Readiness0

Supportability75

Spec posture

Inferred effective spec 2026-07-28+·current baseline 2026-07-28

Static source inference · medium confidence

Current-spec patterns detected

Summary

Scoring was not performed

Issues (7)

CRITICALtool-description ↗

No visible input JSON schemas in source code. Tool definitions show only parameter names ('queryId') without formal type declarations, constraints, or validation rules.

get_photoget_sticker

HIGHtool-description ↗

Descriptions are generic and lack LLM-critical context. 'Retrieve a photo by queryId. Returns the photo from storage, database, or 404 if not found. Serves with appropriate MIME type.' does not explain: What is a queryId? How does the LLM obtain one? What does the response contain beyond the photo? Is this for searching or retrieving a known ID? Descriptions are 129 - 155 chars but lack actionable guidance on when/why to call.

get_photoget_sticker

HIGHtool-description ↗

Parameter 'queryId' is ambiguous. Description says 'The unique identifier for the photo to retrieve' but does not explain: Is this a Telegram message ID? A file ID? A hash? Where does the LLM get this value? Lack of format guidance invites incorrect usage.

get_photoget_sticker

Recommendations

Add formal JSON Schema definitions for all tool inputs. Declare queryId with type 'string', minLength, and a regex pattern or enum if applicable (e.g., 'A Telegram file_id or message_id, 32+ alphanumeric characters').
Rewrite tool descriptions to answer: WHAT does it do? WHEN should the LLM call it? WHAT does it return? Example: 'Retrieve a photo by its Telegram file_id. Returns the image file (JPEG/PNG) plus metadata (dimensions, MIME type, file size). Use after searching for photos with search_photos() to get the file_id. Returns 404 if the file_id is invalid or expired, call search_photos() to find a valid ID.'
Document output schema for get_photo and get_sticker. Specify: Does it return the raw image binary, or a JSON object with image URL + metadata? If metadata, list all fields (mime_type, width, height, file_size, etc.).
Add a search_photos or list_photos tool so LLMs can discover queryIds before calling get_photo. Current tool set is composition-incomplete.
Clarify parameter constraints in descriptions. Example: 'queryId (string, 32+ alphanumeric chars): The Telegram file_id or message_id. Obtain from search_photos() or Telegram client. Invalid IDs return 404.'
Add error recovery guidance. Example: 'If 404: The file_id may be expired. Call search_photos() with the original search terms to obtain a fresh file_id and retry.'
Document health_check response structure. Is it {"status": "ok"} or plain text 'OK' or HTTP 200 with no body?

Score history

Overall score trend

↑ 18 points across a rubric change (v1 → v2)

45/100

Scored	Grade	Overall	Spec posture	Rubric
2026-09-22	F	45	2026-07-28+	v2
2026-03-09	F	27	-	v1