{"slug": "agentstack-mcp-one-deterministic-reasoning-stack-for-ai-agents-simulate-decide", "title": "AgentStack MCP: one deterministic reasoning stack for AI agents (simulate + decide + compute)", "summary": "Developer inity13 released AgentStack MCP, a deterministic MCP server that unifies three previously separate engines for simulation, decision-making, and computation into a single endpoint. The server offers composite tools that chain these engines for end-to-end reasoning, such as evaluating options with scenarios and stress-testing decisions, with a free tier of 20 calls per day.", "body_md": "*The fourth in a suite of deterministic MCP servers for AI agents — and the one that ties the first three together.*\n\nOver the last stretch I shipped three focused, deterministic MCP servers:\n\nThey're great on their own, but agents kept needing all three in the same task — and installing three servers, juggling three keys, and hand-gluing their outputs is friction. So here's **AgentStack MCP**: one endpoint, one key, all three — plus composite tools that chain them.\n\n```\n{ \"mcpServers\": { \"agentstack\": {\n    \"type\": \"http\", \"url\": \"https://agentstack-mcp.pages.dev/mcp\" } } }\n```\n\nFree tier: no key, 20 calls/day. The tools are namespaced so an agent always knows which engine it's calling:\n\n`sim_*`\n\n— ScenarioSim (run, sensitivity, break-even, compare, templates)`decide_*`\n\n— DecisionMatrix (decide, score, sensitivity, compare_two, methods)`calc_*`\n\n— PrecisionCalc (metrics, currency, NPV, IRR, loan, depreciation, …)These chain the engines to do reasoning **no single server can**, deterministically end-to-end:\n\n** evaluate_options_with_scenarios** (simulate → decide) — project each option as its own scenario, then rank the\n\n```\n{ \"name\": \"evaluate_options_with_scenarios\", \"arguments\": {\n  \"template\": \"saas_growth\", \"horizon\": 12,\n  \"options\": [\n    { \"name\": \"Aggressive\", \"inputs\": { \"new_customers_per_period\": 60, \"churn_rate\": 0.05 } },\n    { \"name\": \"Lean\",       \"inputs\": { \"new_customers_per_period\": 20, \"churn_rate\": 0.02 } }\n  ],\n  \"criteria\": [\n    { \"metric\": \"ending_mrr\", \"weight\": 3, \"direction\": \"benefit\" },\n    { \"metric\": \"total_churned_customers\", \"weight\": 1, \"direction\": \"cost\" }\n  ]\n} }\n```\n\n** plan_to_valuation** (simulate → compute) — project a plan, then value its cash-flow line: NPV, IRR, undiscounted total.\n\n** stress_test_decision** (simulate × decide) — stress one scenario assumption across every option and report how often the chosen option survives (robustness) and where it flips.\n\nBundling 24 tools risks drowning an agent's tool-selection. So the endpoint takes a `?profile=`\n\nfilter:\n\n```\nhttps://agentstack-mcp.pages.dev/mcp?profile=finance\n```\n\n`finance`\n\n→ `calc_*`\n\n+ `plan_to_valuation`\n\n`decision`\n\n→ `decide_*`\n\n+ the two decision composites`simulation`\n\n→ `sim_*`\n\n+ all composites`all`\n\n(default) → everything`io.github.inity13/agentstack-mcp`\n\nIf your agents plan, choose, and do the numbers, give them one calculator that does all three — and never drifts. Feedback welcome.", "url": "https://wpnews.pro/news/agentstack-mcp-one-deterministic-reasoning-stack-for-ai-agents-simulate-decide", "canonical_source": "https://dev.to/datanestdigital/agentstack-mcp-one-deterministic-reasoning-stack-for-ai-agents-simulate-decide-compute-597p", "published_at": "2026-08-11 12:30:22+00:00", "updated_at": "2026-08-11 12:49:48.652077+00:00", "lang": "en", "topics": ["ai-agents", "developer-tools", "ai-infrastructure"], "entities": ["AgentStack MCP", "ScenarioSim", "DecisionMatrix", "PrecisionCalc", "inity13"], "alternates": {"html": "https://wpnews.pro/news/agentstack-mcp-one-deterministic-reasoning-stack-for-ai-agents-simulate-decide", "markdown": "https://wpnews.pro/news/agentstack-mcp-one-deterministic-reasoning-stack-for-ai-agents-simulate-decide.md", "text": "https://wpnews.pro/news/agentstack-mcp-one-deterministic-reasoning-stack-for-ai-agents-simulate-decide.txt", "jsonld": "https://wpnews.pro/news/agentstack-mcp-one-deterministic-reasoning-stack-for-ai-agents-simulate-decide.jsonld"}}