MCP servers that ship as code — a repository or a package you install and run locally (usually over stdio). They have no hosted endpoint, so we cannot probe or grade them: instead each one shows its repository, install command, license, stars and when it was last updated.
Collected from GitHub repositories and npm, PyPI, Cargo, Go, Maven, NuGet and RubyGems packages tagged as MCP servers, the official MCP Registry, Docker's MCP catalog, ModelScope, Baidu MCP World, Cline and other public directories; one entry per repository. A server that also offers a hosted endpoint is listed under Remote instead. These entries are not counted in our reachability, verification or readiness statistics.
-
warposgithub.com/jcmrs/warposupdated this year
MCP server for deterministic task execution via Domain Profile Agents + Task Templates. Provides planning/orchestration layer for CLI AI assistants (Warp, etc.)
-
MCP server for DRC (Deterministic Replay Compute) — trace, analyze, search, diff, bisect, replay, and verify captured executions
npm MIT -
Deterministic verification gateway for MCP — risk-gated Python execution in a restricted environment, math/logic/code/SQL verification engines, and skill supply-chain provenance guard
go Apache-2.0 -
Dependency-free Python MCP server that puts a deterministic policy boundary in front of subprocess execution for agents.
MIT -
Deterministic MCP server for governed and auditable AI execution
Apache-2.0 -
Local MCP server for deterministic skill selection. Discovers, validates and plans. Zero dependencies, nothing executed.
pypi Apache-2.0 ★ 1 -
Pre-execution governance for AI agents. Intercepts MCP tool calls before execution with deterministic blocking, human-in-the-loop holds, and behavioral drift detection.
npm MIT -
Auditable deterministic reasoning MCP server (non-LLM). World-graph → A* auditable reasoning → execution → learning loop. Fail-closed: an unchecked claim is never returned as passed. Free for non-commercial agent use.
npm NOASSERTION -
Dynamic analysis for executed Sui Move transactions: invariants over the execution trace, deterministic triage, a budgeted scout, and an MCP server
npm MIT ★ 6 -
aishell-gategithub.com/aishell-gate/aishell-gateupdated this year
AIShell-Gate -- deterministic policy layer for AI generated Unix command execution. Policy gateway for AI agents executing shell commands. MCP server for Claude Code & Cursor. Tamper-evident audit
★ 1 -
Deterministic execution gate, MCP server, and audit ledger for LLM tool calls.
Apache-2.0 -
Deterministic, verified media workflows for software agents. Inspect → plan → execute → verify. Zero silent failures.
npm MIT ★ 3 -
codetopono update data
Domain-blind code graphs for AI agents: symbols, calls, blast radius, proof-checked snapshots. Indexes a repository into a deterministic graph and answers structural questions over MCP — read-only, nothing executed.
smithery -
toolbeltgithub.com/aragaobruno/toolbeltupdated this year
SRE-grade AI Agent Toolchain & Skill Harness. Optimized for context efficiency, robust rate-limiting, and deterministic tool execution.
MIT ★ 2 -
mfg-mcpgithub.com/sbg0700/mfg-mcpupdated this year
AI plans, deterministic code executes, humans approve — on-prem manufacturing data preprocessing via MCP, agents, and a local LLM.
MIT -
Governed expense-operations platform with deterministic policy execution, durable workflows, provenance, evaluations, observability, and MCP interfaces.
MIT -
Household permission layer for simulated Alexa+ experiences: MCP actions, role-aware approvals, deterministic execution, and auditable receipts.
MIT -
Secure AI agent execution with MCP, deterministic policies, human approval, idempotency and auditable simulated financial operations.
MIT -
timeseries-mcpgithub.com/Lkhanaajav/timeseries-mcpupdated this year
Deterministic time-series statistics for AI agents — typed MCP tools for anomaly detection, changepoints, decomposition, trend tests, and data-quality auditing. No code execution.
MIT -
DecompGTIgithub.com/yuvraj-rathod-1202/DecompGTIupdated this year
DecompGTI is a graph-reasoning framework that fine-tunes small language models to parse natural-language graph queries into structured JSON plans executed by deterministic algorithms on a FastMCP server.
MIT -
Guarded AI-assisted golf trade-in workflow with deterministic orchestration, human review, RAG, MCP-compatible tools, and auditable model execution.
MIT -
cwgithub.com/Queopius/cwactive
Governed AI-assisted software development with Codex: phased execution, deterministic validation, independent reviews, and evidence-based approval gates.
Apache-2.0 ★ 2 -
Local-first TouchDesigner AI copilot with 114 MCP tools, truthful concept routing, deterministic visual techniques, transactional execution, live validation, and rollback.
npm MIT -
Deterministic pre-execution safety gate for AI agents: verify tool calls against trusted evidence, policies, and WorldState before side effects.
★ 1 -
Verify AI-generated code with execution evidence — deterministic, local-first verification for coding-agent changes via a macOS app, CLI, and MCP server.
MIT ★ 1 -
MCP server and agentic control layer for KiCad 10, combining LLM-based planning where useful with deterministic execution, state tracking, evidence, rollback, and KiCad validation.
AGPL-3.0 ★ 15 -
The execution layer for AI tax prep — agents think, taxfill fills, verifies, and gets it mailed. Guided intake + deterministic fill/verify/calc tools + versioned form & jurisdiction knowledge, for any MCP client.
NOASSERTION ★ 5 -
Native Android device-control and automation platform: lets authorized AI agents operate your own phone over MCP, with an on-device policy engine, approvals, Shizuku-backed shell execution, and deterministic skills.
MIT -
Turns vague requests into structured prompts, execution plans, and policy-aware workflows — with readiness verdicts, PR merge-safety checks, agent packs, and MCP tool exports. Offline-first, deterministic, provider-agnostic.
Apache-2.0 ★ 29 -
⚡ High-frequency (~60Hz) in-browser behavior tree execution engine for autonomous AI agents. Reactive priority interrupts, safety guardrails & deterministic action replay over MCP.
npm MIT ★ 8 -
Permission-aware execution layer for AI agents: a self-hosted MCP server where the model proposes, deterministic policy decides, and the owner approves. Simulated Alexa+ home, hash-chained audit.
MIT -
pyvisualizergithub.com/haider1998/pyvisualizerupdated this year
A deterministic, AST-verified call graph for any Python project — every edge traceable to a file:line. The reproducible, CI-gateable alternative to LLM-generated diagrams. No code executed. Nothing leaves your machine.
MIT ★ 12 -
Kern , The local, deterministic code-intelligence engine for AI agents. Fast AST indexing, precise dependency graphs, sandboxed execution, and MCP tools with zero network calls and zero telemetry.
go MIT ★ 1 -
Deterministic, non-executing evidence and change monitoring for public npm MCP servers. Reports what changed since the version you approved. It never claims a package is safe.
npm Apache-2.0 ★ 1 -
NL→SQL agent over a lakehouse with five deterministic control rings: AST allowlist guard, cost budget estimated before execution, role-based masking and a tamper-evident audit chain. Served over MCP.
Apache-2.0 -
Institutional-grade autonomous trading agent decoupling Google Gemini reasoning from a 100% deterministic Python Hard Risk Gate. Executes regime-adaptive, defined-risk options strategies via Alpaca API & FastMCP.
MIT -
Preflight Checks (x402)no update data
Eight deterministic pre-execution verification tools for AI agents (npm/PyPI/crates slopsquatting check, exact version & hash check, EVM tx/calldata byte decode, Bitcoin raw tx/PSBT decode, Bitcoin UTXO chain context, EUR-Lex legal citation check, CVE lookup, EU AI Act obligations retrieval). Paid per call via x402 USDC on Base.
smithery -
Open-source MCP server that gives AI agents the QA lead’s job: test strategy, quality gates, test generation and execution (Playwright, Cypress), flaky triage, release readiness. Judgments with evidence, deterministic guardrails, a replayable trace. Capabilities first, frameworks later.
MIT -
Validated analytics agent for CSV/Excel. LLMs plan - as a JSON plan or a bounded tool-calling loop - while deterministic pandas/scipy/DuckDB computes and re-verifies every number. No arbitrary code execution: the model only selects audited operations.
MIT ★ 4 -
nexus-mcp-extensiongithub.com/tanaikech/nexus-mcp-extensionupdated this year
Nexus-MCP solves Tool Space Interference in LLMs. It uses a single gateway to aggregate multiple MCP servers and employs a deterministic 4-phase workflow (Discovery, Mapping, Schema Verification, Bridged Execution) to prevent context saturation and tool hallucinations, maintaining reasoning accuracy with massive tool ecosystems.
MIT ★ 4 -
SpendGuard AI: a production-shaped MCP portfolio monorepo for FinOps. Investigate cloud-spend anomalies, draft a savings action, obtain independent approval, and execute it through a deterministic simulated provider that never touches a real cloud account.
-
neuro-mcp-unified-orchestration-enginegithub.com/bitsandbrainsai/neuro-mcp-unified-orchestration-engineupdated this year
Production-grade NeuroMCP orchestration engine that unifies n8n workflows with agentic LLM ecosystems, delivering secure, deterministic, and scalable automation. Features context-aware execution, multi-transport support, structured observability, & fault-tolerant design, powering AI-native, resilient, & high-performance enterprise infrastructure.
-
A CLI-first autonomous coding agent where the actor never grades itself. Every edit is grounded in your repo's real topology and verified by an independent deterministic grader (drift, lint, verify-task, intent) before it ships. Closed-loop: comprehend → plan → execute → critique → replan. Also a passive MCP harness.
go Apache-2.0 ★ 25 -
reasonforge-publicgithub.com/euuuuuuan/reasonforge-publicupdated this year
A deterministic, read-only reasoning harness that a coding agent can call over 8 MCP tools: it forces a scope lock, a plan whose first move is the cheapest way to be proven wrong, a blind-spot sweep with line-level evidence, and a verdict gate that refuses a verified-fact claim without proof of execution. Read-only is enforced by a test.
Apache-2.0 -
AI-Atelier, the all in one personal "atelier" (means handcraft studio in french) that can adapt to your need.
npm MIT ★ 4 -
Treat normalized scan findings as immune signals: discover patterns, synthesize non-executing defenses, remember what worked.
MIT -
One auditable remediation transaction: frozen pre-state, policy gate, execution receipts, verified post-state, rollback on failure.
MIT -
Agent Behaviour Drift Sentinel: detects silent behavioural drift in agent execution traces against a signed baseline. Reference implementation of the public SHPBL behaviour specification.
MIT ★ 1 -
Agent action evidence ledger: append-only, hash-linked record of agent proposals, executions, redactions and annotations with an independent verifier. Reference implementation of the public SHPBL behaviour specification.
MIT -
Exact math execution layer for AI agents: fractions, finance, units, dates. MCP server + DSH plugin, 19 packs, fail-loud errors, 19,000 acceptance tests.
MIT -
Kine(/kaɪn/)· 为 AI 智能体设计的确定性视频剪辑/合成引擎:第 N 帧 = f(N)。TypeScript monorepo(8 包)+ Edge CDP 逐帧抓帧 + ffmpeg 分片编码 + Python 侧车。能力:32 个 CLI/MCP 工具,读事实→改 DSL→量结果→验证出片;声明式 timeline.yaml 单一真源;逐帧确定性渲染、字节级复现;8 类指标 + 断言集 + 基线回归;34 个 widget;15 种转场;离线 SFX;字幕;浅色模式。已出片 world.execute(me) MV(25 幕 / 121.9s)等。
MIT ★ 1