A Telegram bot that provides live CS2 FACEIT data including player stats, match history, player comparisons, leaderboard, and ELO trends. Includes an MCP server exposing FACEIT tools for Claude Desktop integration.
Five tools with clear verb-noun naming (get_*, compare_*) and documented descriptions (100-150 chars each). All tools have input schemas with type definitions and parameter descriptions. However, output schemas are not formally documented, responses are JSON strings without declared field structures. Error handling returns JSON with 'error' keys but lacks recovery guidance. No tool annotations (readOnlyHint, etc.). Parameter descriptions are adequate but lack format constraints (e.g., 'limit: 1 - 20' stated in description but not as JSON Schema min/max). Missing documentation of response field structures forces LLMs to parse unstructured JSON.
Compare CS2 FACEIT stats for 2–6 players side by side.
Return stored ELO snapshots for a registered nickname, showing ELO progression over time.
Return all bot-registered users ranked by their current live FACEIT CS2 ELO.
Return the most recent CS2 matches for a FACEIT player.
Return ELO, skill level, region, and lifetime CS2 stats for a FACEIT player.
Output schemas not documented. Tools return JSON strings without declared field structures (e.g., get_player_stats returns nested 'stats' object with 'kd', 'hs_pct', etc., but no schema definition). LLMs must infer response structure from examples, increasing parsing errors and token waste.
Parameter constraints not formalized in JSON Schema. 'limit' parameter in get_match_history states '1 - 20' in description but lacks minItems/maxItems in schema. Numeric bounds should be enforced at schema level, not just documented.
Error responses lack recovery guidance. Errors return {'error': 'Player not found'} without suggesting next steps (e.g., 'Try search_players() with a partial name'). Pattern:recovery-guide requires actionable error messages.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | D | 55 | 2026-07-28+ | v2 |
No tool annotations. Tools lack readOnlyHint, destructiveHint, or idempotentHint metadata. All tools are read-only (safe to retry), but this is not declared in the schema, forcing LLMs to infer safety.
Pagination not implemented. get_leaderboard and get_match_history return unbounded lists without limit/offset or cursor support. Large result sets risk context window exhaustion.