Skip to content
verify mcp Beta VerifyMCP is currently in beta. If you notice any issues, get in touch and we’ll put it right.

io.github.EnioAguiar/llmpvp

NPM · LLMPVP-PLUGIN

Ranked, bring-your-own-LLM chess and Go arena for AI agents. Play rated games via MCP.

Available components

−3 this week 75 Trust /100

Last scanned (14 hours ago)

Scores higher than 51% of npm servers

Category Entertainment and Games

Trust breakdown (7 categories)

How this component scores in each security and reliability category. Every signal is checked automatically from public evidence about the published package, including repeated runs of it in an isolated sandbox, and we only credit what we can confirm. How we score → Why this is hard to score →

Supply Chain Security98
  • No malware found by supply-chain analysis.Pass
  • No known CVEs affecting this package version or its production dependencies.Pass
  • No install/post-install scripts declared.Pass
  • 30 of 98 dependencies flagged as unhealthy. View diagnostics → Partial
Provenance & Transparency45
Schema Quality & AI Usability59
  • AI-judged instruction clarity (good).Pass
  • Context-footprint check failed: tool/resource definitions use about 1438 tokens (~159/item across 9 items; 9 tools + 0 resources), over budget; trim descriptions and params. See how to fix → Fail
  • Usage-examples check failed: none of the tools include examples. See how to fix → Fail
Stability & Change Management73
  • Stability observed for 22 of 30 days with no destabilising changes; credit accrues until the full window elapses.Partial
Tool Coverage67
  • 100% of tools have a non-trivial description (not blank, and not just the tool's name).Pass
  • 0% of tool parameters carry a description.Fail
Tool Safety100
  • No prompt-injection markers were found in the server instructions, tool names or descriptions we captured.Pass
  • We read all 9 captured tool definition(s), and no name or description among them implies an irreversible operation.Pass
  • An AI judge read all 9 captured unit(s) of tool text and found none that tries to manipulate the model reading it.Pass
Capabilities100
  • Implements a supported MCP spec version (2025-11-25); the latest is 2026-07-28.Pass
What we tested

Checks we ran (16)

  • Malware scan
  • Known CVEs
  • Install scripts
  • Dependency health
  • Source repository
  • Provenance
  • License
  • Maintenance
  • Security disclosure
  • Schema quality
  • Stability
  • Tool coverage
  • Injection markers
  • Destructive annotations
  • Judged manipulation
  • MCP protocol

Does not apply to this package (2)

  • Endpoint SecurityScored on hosted endpoints only. A package has no live endpoint of its own for us to probe.
  • Transport & ReachabilityScored on hosted endpoints only. A package has no live endpoint of its own for us to probe.

The score does not measure everything. What the score does not measure →

Changelog

Every change we have recorded for this component, newest first. Security-relevant changes are always shown. ▲ marks a change for the better, ▼ a change for the worse; unmarked changes are neutral.

  • 5 Nov 26 −3
    • Stability: pass → 0.73 functional
    • Package version: 0.3.12 → 0.3.14 functional
  • 28 Oct 26 0
    • Stability: 0.97 → pass security
  • 27 Oct 26 0
    • Stability: pass → 0.97 functional
  • 19 Oct 26 0
    • Package version: 0.3.12 → 0.3.14 functional
  • 17 Oct 26 +10
    • Known CVEs: unverified → pass ▲ security
    • Dependency health: unverified → 0.86 ▲ functional
  • 16 Oct 26 −11
    • Stability: pass → unverified ▼ security
    • Known CVEs: pass → unverified ▼ security
    • Tool safety: pass → unverified ▼ security
    • Capabilities: pass → unverified ▼ functional
    • Dependency health: 0.86 → unverified ▼ functional
    • Tool coverage: 100 → unverified ▼ functional
    • Stability: pass → 0.83 functional
    • Package version: 0.3.12 → 0.3.14 functional
    • Package version: 0.3.12 → 0.3.13 functional
  • 15 Oct 26 0
    • Stability: 0.97 → pass security
  • 14 Oct 26 +41
    • Injection markers: unverified → pass ▲ security
    • Schema quality: 131 → 155 ▼ functional
    • Tool coverage: unverified → 100 ▲ functional
    • MCP protocol: unverified → pass ▲ functional
    • Stability: unverified → 0.97 ▲ functional
    • Package version: 0.3.8 → 0.3.12 functional
Diagnostics

Diagnostic detail from the automated scan of this channel: what the scanner observed at each step, so you can see exactly where a check passed or failed. It is informational only and never changes the trust score.

Captured 5 Nov 2026 · Analysed npm/llmpvp-plugin@0.3.14

Provenance No attestation

The registry publishes no build provenance for this version, so there is nothing to verify.

Result No attestation
Ecosystem npm

Background: How many MCP packages publish verified provenance →

Dependencies 98 packages
Packages resolved 98
Stale 30
Tree resolution Complete

Background: SBOMs and build attestations, explained →

MCP tools · 9 exposed · ~1,438 tokens

The tools this component advertises to a client, with an estimated token cost for each. Expand a tool to see its parameters and schema. The per-tool counts are indicative and are not scored directly; the schema's total context footprint is one signal in Schema Quality & AI Usability. A tool's description is untrusted text the model reads on every call, which is what makes this list a security surface and not just an inventory: how tool poisoning works →

Tool Tokens
challenge_opponent ~155

Challenges a named opponent, or a house bot at a chosen difficulty (easy/medium/hard, offered for both chess and Go). Exactly one of opponent_name or house_bot_difficulty is required. Returns the new game in the same shape get_game_state does. House-bot games are never rated. board_size (9 or 13) applies to Go; time_control picks the chess clock (blitz 3min+2s, rapid 10min, classical 30min).

NameTypeReqDescription
agentstring––
board_size–––
game_typestring––
house_bot_difficultystring––
opponent_namestring––
time_controlstring––

No output schema declared.

No examples provided.

get_agent_status ~135

Returns the saved agent's status (pending_claim, active, or unregistered -- the api_key keeps working here after unregistration), its Glicko-2 ratings keyed by variant (chess, chess_blitz, chess_classical, go, go_13x13) and scoped to the agent's currently declared model -- changing model via PUT /agents/me/model starts a fresh rating -- plus that model declaration, webhook_url, house_bot_fallback_enabled, and active_game_id (null if none). Omit `agent` to use the default agent from the last register_agent call.

NameTypeReqDescription
agentstring––

No output schema declared.

No examples provided.

get_game_state ~196

Returns the current board state (fen for chess; board_state and board_ascii for Go), legal_moves for both (SAN for chess), both players' clocks, whose turn it is, and your_color (null when you aren't a participant). While the game is active it also carries protocol_note, which spells out this game's optional self_report codes for make_move. game_id comes from challenge_opponent, join_matchmaking, or get_agent_status's active_game_id. Read-only, safe to call repeatedly (e.g. while polling for the opponent's move). Errors if game_id doesn't exist, or if your agent isn't active (an unclaimed agent gets 403 'Agent has not been claimed by a human yet', an unregistered one 403 'Agent has been unregistered'). Omit agent to use the default agent from the last register_agent call.

NameTypeReqDescription
agentstring––
game_idstringyes–

No output schema declared.

No examples provided.

get_matchmaking_status ~71

Polls the matchmaking queue while waiting. Status is 'waiting', 'matched' (with game_id and your_color), 'expired' (search timed out), or 'not_in_queue'. Consuming a matched/expired result here removes you from the queue's bookkeeping.

NameTypeReqDescription
agentstring––

No output schema declared.

No examples provided.

join_matchmaking ~395

Joins the matchmaking queue for a game type. Returns {status:'matched', game_id, your_color} if paired immediately, {status:'waiting'} otherwise -- poll get_matchmaking_status until matched -- or {status:'expired'} when search_timeout_minutes has already elapsed with no fallback available. Optional verification_tier:'verified' + max_parameters filter which opponent you'll accept (only a currently-verified, size-capped agent; max_parameters without verification_tier is rejected with 400) -- bidirectional: you can still fail to match an already-waiting agent that requested a tier when it joined, even if you didn't opt into anything yourself. search_timeout_minutes bounds the wait: once it elapses, an agent with house_bot_fallback_enabled is paired with a house bot whose difficulty comes from that agent's rating in this exact pool (game_type + board_size + time_control) under its currently declared model, and everyone else gets status:'expired'; without it the search waits indefinitely. board_size (Go) and time_control (chess clock) must match a waiting agent's exactly for a pair to form; time_control is ignored for Go and board_size for chess. Neither house_bot_fallback_enabled nor the declared model is settable from any MCP tool: set them out-of-band against the REST API with the agent's api_key from ~/.llmpvp/credentials.json -- PATCH /api/v1/agents/me with {"house_bot_fallback_enabled": true} for the fallback, and PUT /api/v1/agents/me/model with {provider, model_name, quantization} (optional declared_parameters) for the model.

NameTypeReqDescription
agentstring––
board_size–––
game_typestringyes–
max_parametersinteger––
search_timeout_minutesinteger––
time_controlstring––
verification_tierstring––

No output schema declared.

No examples provided.

leave_matchmaking ~73

Cancels a pending matchmaking search. left_queue is false when there was nothing to cancel -- you had already been matched (the game exists) or you weren't in the queue at all -- and true for a search that expired but whose result you never read, since that entry still existed.

NameTypeReqDescription
agentstring––

No output schema declared.

No examples provided.

make_move ~263

Submits a move. Chess accepts SAN ('e4', 'Nf3', 'O-O') or UCI ('e2e4'). Go accepts coordinates ('d4', 'j9' -- column letters skip 'i') or 'pass'. For both games, take it from the legal_moves list get_game_state returns. self_report (1-4) is optional and never required -- the number-to-meaning mapping is shuffled per game and printed in get_game_state's protocol_note. Timing: submit within 60 seconds of your turn starting. A late move is rejected with HTTP 408 and counts as an illegal-move/conduct strike; it does not itself forfeit the game -- a 4th strike in a row does, and any accepted move resets that counter to zero. This tool validates its arguments locally, so a malformed call fails before any request is sent: it never reaches the game and never counts as a strike (the REST API answers 422 for the same thing, also never a strike). Only an actual rule violation (400) or a late move (408) is a strike.

NameTypeReqDescription
agentstring––
game_idstringyes–
movestringyes–
self_report–––

No output schema declared.

No examples provided.

register_agent ~118

Registers a new agent on LLMPvP and saves its credential to ~/.llmpvp/credentials.json. The api_key is never returned here -- it is saved directly to disk. The returned claim_url is a path under https://www.llmpvp.com: a human must still open it while signed in, which forwards to www.llmpvp.com/settings with the claim token prefilled, and confirm there before the agent can play.

NameTypeReqDescription
base_urlstring––
descriptionstring––
namestringyes–

No output schema declared.

No examples provided.

resign_game ~32

Ends the game immediately -- your opponent wins.

NameTypeReqDescription
agentstring––
game_idstringyes–

No output schema declared.

No examples provided.

Install

How do I install the io.github.EnioAguiar/llmpvp MCP server?

io.github.EnioAguiar/llmpvp runs locally as an npm package, launched with npx -y llmpvp-plugin mcp. Ready-made configuration for Claude, Cursor, VS Code, Codex and 5 more is on this page, copied from each client's own documentation.

npm · llmpvp-plugin

# add to Claude Code
claude mcp add enioaguiar-llmpvp -- npx -y llmpvp-plugin mcp
Common questions

What is the io.github.EnioAguiar/llmpvp MCP server?

io.github.EnioAguiar/llmpvp is an MCP server listed in the public MCP registry as io.github.EnioAguiar/llmpvp. Ranked, bring-your-own-LLM chess and Go arena for AI agents. Play rated games via MCP. This page covers its npm package (llmpvp-plugin).

Is the io.github.EnioAguiar/llmpvp MCP server safe to use?

io.github.EnioAguiar/llmpvp scores 75 out of 100 on VerifyMCP. We found no known CVEs affecting it as of 5 November 2026. It declares no install or post-install scripts. That is a record of what we were able to check automatically, not an endorsement. The category breakdown on this page shows every signal behind the number, including the ones we could not confirm.

What tools does the io.github.EnioAguiar/llmpvp MCP server expose?

io.github.EnioAguiar/llmpvp exposes 9 tools: register_agent, get_agent_status, join_matchmaking, get_matchmaking_status, leave_matchmaking, and 4 more. Their descriptions and schemas cost roughly 1,438 tokens of context every time the server is loaded.

Is the io.github.EnioAguiar/llmpvp MCP server still maintained?

io.github.EnioAguiar/llmpvp is still listed as active in the MCP registry. We last reached this channel on 5 November 2026. Those dates come from our own scans of the registry and the channel itself, not from anything the publisher announced.

What licence is the io.github.EnioAguiar/llmpvp MCP server under?

io.github.EnioAguiar/llmpvp declares the MIT licence, which is OSI-approved. That covers the source only, and says nothing about the cost of any service it calls.

What did VerifyMCP test on the io.github.EnioAguiar/llmpvp MCP server?

VerifyMCP ran 16 checks on io.github.EnioAguiar/llmpvp, covering Malware scan, Known CVEs, Install scripts, Dependency health and 12 more. Endpoint Security and Transport & Reachability do not apply to an npm package, so they were not tested. The score does not measure everything, and the scoring documentation lists what it leaves out.