"author:itsryanarmani" matching MCP servers:

Enables agents to list and retrieve pizza support policy tasks, run deterministic benchmark evaluations, score submitted answers, and inspect provenance through MCP tools. It supports offline and provider-backed runs for testing refund, escalation, and tool-use behavior.

MCP server for auditing agent-native readiness of a product directory or URL and generating artifacts like AGENTS.md and server.json.

Enables starting and monitoring shared multiplayer sessions for agentic coding terminals, allowing agents to host wrapped commands and retrieve session status via WebSocket.

Enables agents to list, read, mark, send, and manage notifications and preferences through MCP tools.

Exposes CRM truthfulness evaluation tasks so agents can list and retrieve scenarios, score submitted answers against consent and claim-ledger rules, run benchmarks, and fetch dataset provenance. Enables programmatic auditing of whether a sales agent claims only what its evidence supports.

Exposes MCP tools to list and inspect the 48 harm-ladder tasks, deterministically score a submitted decision trace, run a full evaluation against a mock or real model provider, and retrieve data provenance. This lets developers measure a model's pushback and refusal calibration as plan harm rises, directly from any MCP client.

Exposes five tools that let an assistant list and inspect schema-mutation benchmark tasks, grade submitted JSON tool calls against canonical answers, run the benchmark against a chosen model, and read dataset provenance. It enables measuring whether a model follows the current tool schema or falls back on a memorized one.

Enables agents to browse and retrieve benchmark reasoning tasks, deterministically grade submitted answers, run the evaluation suite against a model provider, and inspect data provenance and the ranking-stability index entry. It also supports scoring stability analysis across repeated runs and prompt variants via its tools.

Enables users to list and inspect benchmark tasks, grade answers, run benchmark evaluations, and retrieve provenance through its MCP interface.

Enables users to list and retrieve evaluation tasks, score recorded submissions, run benchmark evaluations via OpenRouter, and access provenance information through MCP tools.

Enables agents to share live terminal sessions, manage viewers and join requests, and stop active shares, with consent handled through tool-approval prompts.

Enables MCP hosts to access the user's local token-usage league profile and matching candidates in the same or adjacent tiers, while keeping raw usage on-device.

Enables listing and retrieving tasks from a benchmark for rule-conflict scenarios, scoring model answers, running evaluations, and checking data provenance through MCP tools. It supports offline mock runs and OpenRouter-backed model evaluations with deterministic assertion grading.

Enables MCP clients to list and inspect coding-agent benchmark tasks, score recorded answers, run evaluations, and retrieve data provenance.

Lets MCP clients browse benchmark tasks, inspect individual tasks, grade answers, execute the benchmark suite, and fetch provenance metadata.

Enables peer-to-peer AI inference donation by exposing MCP tools to check node status and request capacity, with local-first compute sharing, consent, and metering.

Enables MCP clients to list and inspect evaluation tasks, score model answers, run the full benchmark, and retrieve provenance information.

Enables MCP clients to list and inspect benchmark tasks, score submitted answers, run evaluations through OpenRouter, and retrieve provenance data for exploit-resistance testing.

Enables clients to list and retrieve skateboard-physics benchmark tasks, score submitted answers deterministically, run the benchmark, and inspect provenance via MCP tools.

Enables users to list and inspect household memory-conflict tasks, score answers, run evaluations, and retrieve provenance data through MCP tools.