Announcing VLRdevAPI v3.1: MCP Server, Universal Tabular Exports, and Python 3.15 Ready

Native Model Context Protocol (MCP) ServerCopy Anchor Link

AI coding assistants and autonomous agents can now inspect Valorant esports data natively. VLRdevAPI v3.1 includes a production-grade MCP server supporting the standard Model Context Protocol.

Run the server directly via the CLI:

vlrdevapi-mcp

Or run without prior installation via uvx:

uvx --from "vlrdevapi[mcp]" vlrdevapi-mcp

Configured for Claude Desktop & CursorCopy Anchor Link

Connect Claude Desktop by adding vlrdevapi to your claude_desktop_config.json:

{
  "mcpServers": {
    "vlrdevapi": {
      "command": "uvx",
      "args": ["--from", "vlrdevapi[mcp]", "vlrdevapi-mcp"]
    }
  }
}

40 Tools + Specialized Composite Analysis ToolsCopy Anchor Link

The MCP server provides 40 tools spanning matches, series, teams, players, tournaments, and editorial news. It also introduces high-level composite tools designed for complex reasoning tasks:

• get_match_full_context(series_id): Combines series overview, player ratings, and VODs in a single call.
• get_team_overview(team_id): Gathers team profile, active roster, recent match results, upcoming fixtures, and tournament placements.
• compare_teams(team_id_1, team_id_2): Compares rosters, rankings, earnings, and recent form head-to-head.
• get_player_career_summary(player_id): Consolidates player biography, team history, agent ratings, and tournament placements.
• get_h2h_matchup(team_id_1, team_id_2): Extracts historical match records between two teams.
• get_tournament_dossier(event_id): Overview of tournament stages, participating teams, and final standings.
• get_live_and_upcoming_schedule(limit): Merges the live match ticker with upcoming schedules.
• get_news_digest(limit): Editorial digest with optional full Markdown article body.
• series_map_summary(series_id, game_id): Map-level summary combining economy, replay timeline, and player stats.

All tools trap domain errors gracefully and return structured JSON responses, ensuring your AI assistants never crash during interactive sessions.

Universal Tabular Export LayerCopy Anchor Link

Transforming Valorant esports data into dataframes and spreadsheets previously required manual dictionary comprehension loops. In v3.1, every model inherits from ExportableModel:

import vlrdevapi

# Fetch series player statistics
players = vlrdevapi.series(754737).players()

# 1. Convert to pandas DataFrame
df = players.to_df()
print(df[["name", "team_name", "rating", "acs", "kills", "deaths"]].head())

# 2. Convert to native Polars DataFrame (with typed nested list columns)
polars_df = players.to_polars()

# 3. Export to CSV (with semicolon-delimited multi-value lists)
players.to_csv("match_stats.csv")

# 4. Export to Apache Parquet
players.to_parquet("match_stats.parquet")

# 5. Export to Excel (.xlsx)
players.to_excel("match_stats.xlsx", sheet_name="GrandFinal")

# 6. Export to in-memory JSON or file
json_str = players.to_json(indent=2)

# 7. Generate a GitHub-flavored Markdown table
print(players.to_markdown())

Thoughtful Semantic FlatteningCopy Anchor Link

Our tabular extraction system is intentionally designed for analytics:

• No static columns: Upcoming and completed match tables omit meaningless static status columns.
• TBD matchups resolved: Matches without determined opponents render "TBD" instead of empty gaps.
• Map segregation: Multi-map series player stats distinguish individual maps from aggregate views ("All Maps").
• Spreadsheet-safe lists: Multi-value fields (e.g. agent pools or roster roles) use standard semicolons ("Player;IGL" or "neon;raze;jett"), preventing malformed CSV formatting.

Modular Dependency ExtrasCopy Anchor Link

Data science libraries are heavy. To keep vlrdevapi lightweight for bots and microservices, export dependencies are strictly optional:

• pip install "vlrdevapi[pandas]"
• pip install "vlrdevapi[polars]"
• pip install "vlrdevapi[excel]"
• pip install "vlrdevapi[parquet]"
• pip install "vlrdevapi[mcp]"
• pip install "vlrdevapi[all]"

Calling .to_df() without pandas installed will raise a descriptive MissingDependencyError with the exact pip install command.

Python 3.15 Ready & Production StableCopy Anchor Link

VLRdevAPI is tested against Python 3.11, 3.12, 3.13, 3.14, and the upcoming Python 3.15 nightly builds. With 1,164 automated tests covering every parser, scraper route, export format, and MCP tool, v3.1 is ready for production workloads.