"Actual Budget" matching MCP servers:

Model-agnostic code memory MCP server using budget-aware graph retrieval for AI agents, providing sub-millisecond queries, token budgeting, and deterministic results without embeddings or vector databases.

Provides tools for Bazi (Chinese astrology) chart calculation and analysis, enabling LLMs to generate accurate birth charts, determine patterns, and answer follow-up questions based on actual calculations rather than model knowledge.

Persistent memory engine for AI coding agents. Single Go binary, zero runtime dependencies, MCP-native. Stores, searches, and deduplicates memories across sessions using embedded SQLite with hybrid FTS + semantic search, memory decay, relation graph, and token-budget context assembly.

An intelligent, agentic MCP server gateway to OpenRouter's 200+ AI models. It provides budget controls, security features, semantic memory, and code search capabilities.

TinyContext provides a local, token-efficient memory layer for AI agents. It enables saving and recalling memories via MCP tools, using hybrid BM25 and dense retrieval to return only context that fits the requested token budget.

Provides persistent memory for coding agents and grounds their claims by verifying against the actual codebase, preventing hallucinated responses.

Enables MCP hosts to query and maintain a shared, local-first long-term memory stored as plain markdown files, with budget-aware recall and gated writes. It supports auditable recall and write receipts, failing closed when writes violate trust, deduplication, or destructive-change rules.

Gives AI agents persistent, shared project memory so they can record decisions, failed attempts, tasks and notes, then retrieve a ranked, budget-trimmed slice of what matters across sessions and tools. Access is metered per call in USDC over x402 with no account or API key required.

An MCP-native agentic platform orchestrating planner/executor/critic agents over hybrid RAG with three-tier memory, budget enforcement, safety guardrails, and full observability. It exposes all capabilities as MCP tools, enabling natural-language control of document ingestion, retrieval-augmented generation, and multi-step AI workflows.

An MCP server that transforms Claude into an honest CV co-pilot, mechanically verifying all claims against your actual experience through a structured truth base.

Enables coding agents to access structured product context (decisions, goals, evidence) from meetings and tools, ensuring they build from actual product decisions.

Enables AI agents to compress and selectively retrieve context, with measured recall rather than claimed performance. It provides tools to assess potential traffic and token savings, list compression dictionaries, and assemble relevant memory entries within a token budget.

Indexes a codebase into a symbol-level graph and exposes tools for finding symbols, querying relationships, and assessing impact, letting AI coding agents answer structural questions in a single call within a token budget.

Saves the current state of a project — decisions, rejected approaches, open questions and the actual files — and hands it to any new chat, so you never re-explain a project twice. Local-first: everything is written to your own disk as plain Markdown and JSON, with no account and no telemetry.

MCP-native persistent-memory agent that remembers user preferences across sessions, forgets superseded facts, and recalls relevant memories within a tight token budget.

Filters an agent's already-retrieved memories to fit a token budget using the memory_select tool, while keeping API keys in the environment. It works locally via MCP stdio or HTTP with optional JEV gateway support.

Cost-first memory layer for MCP-capable agents that stores memories cheaply and recalls relevant slices under a hard token budget, logging receipts for every recall.

Exposes read-only capabilities from the Baihua home server to any MCP client via stdio, including knowledge base search/reading, budget summary, and task listing.

Turns Confluence documentation into an AI-powered knowledge base, enabling natural language questions about your systems with answers retrieved from your actual documentation through semantic search.

Marrow is an MCP server that solves agent amnesia. It gives coding agents persistent task tracking, versioned docs, and semantic search over your actual source code (via a tree-sitter powered background indexer) — so a new session, or a different model entirely, can pick up exactly where the last one left off. Built for multi-agent, multi-session software delivery.