Archstone
Local installio.github.Archstone-Romania/archstone
Compile your own business capabilities from declarative YAML and serve them as MCP tools.
4,941 servers in AI & Agents
Compile your own business capabilities from declarative YAML and serve them as MCP tools.
Second opinion for agents: 3-4 AI models answer, one verdict with confidence and dissent.
The fix desk for AI agents: human fixes become skills the whole fleet reuses.
Check an agent work row, post one, and preflight a no-spend cap before any spend/send.
Read your recorded AI agent runs, and what changed when you swapped a model or prompt.
Connects the AI agents you use: your agent hands a task to another and gets the answer back
Design transformer LLM architectures and report their parameters, FLOPs, memory and cost
Verifiable decision records for AI agents: signed receipts for calls to its own governed tools.
Local-first AI memory MCP server. Record, recall and consolidate memory on your own machine.
Persistent memory for AI agents — 36 cited neuroscience mechanisms, local-first, hybrid retrieval.
Persistent agent memory with a decision layer: replay, restore, verify, or none — explainable.
68k+ security-audited MCP skills + free agent trust layer: signed ATC cards, tool gating, audits.
MCP proxy that checks every tools/call against Agent Trust policy, with evidence, before it runs.
Read-only tools to inspect your agent setup: list MCP servers, flag unpinned ones, test a policy.
A social network for AI agents: paint a daily self-portrait and talk with the humans who reply.
Task queue, decisions and memory for coding agents. Local SQLite, no network. Requires Bun.
Optimizing compiler for agent tools: many MCP servers or REST APIs in, one lean endpoint out.
Other AI CLIs (Gemini, GPT, Claude, opencode, Ollama) as a council — ban-safe, no API keys.
Work-history network for AI agents: register, find agents by skill, consult a free house C-suite.
Local-first memory for AI coding agents: shell history with exit codes, git diffs, docs, MCP.
Work-history network for AI agents: register, find agents by skill, consult a free house C-suite.
Measures the cost, tokens and duration of a completed agent workflow from its session log.
One install that runs every sibling MCP server as a child process and merges their tools behind one
Admission gate for AI agents: submit a step, get admit/hold/reject vs your constitution.
Watch video and live sessions, keep timestamped evidence, and verify an agent's own work.
Persistent memory and context for your agent, on a board you can see. Capped, so it stays small.
Local-first agent memory over SQLite: remember, recall, link, forget. Degrades, never fails.
Lessons other agents already learned, by stack and situation. Read before a task, leave one after.
Private local-first AI workspace for shared context, knowledge, tasks, and memories
Context selection and compression for AI agents - recall measured, not claimed.