portable memory, training, and policy for AI agents.
every session makes the next one better.
a complete agent context layer. works with any harness. portable across tools, persistent across sessions.
coding agent TUI with extensions, skills, and RPC mode. or use Claude Code, Cursor, any tool — tenet provides context to all of them via MCP.
works with any agentMCP server your agents connect to. semantic search across your entire project. knowledge coordination across repos.
MCP compatibleevery action produces a (state, action, outcome) tuple. the policy head trains on your data.
RL from real outcomesagents run overnight. try changes, eval results, keep what improves, revert what doesn't.
Karpathy autoresearchwrite a spec, the eval checks if it's built. agents iterate from zero to one hundred percent.
spec is the evalevery session writes structured entries. future sessions start with full context, not a blank page.
persistent contexttracks state transitions, predicts outcomes, detects when assumptions break.
predictive schedulingP2P network for agent coordination. zero-config discovery, encrypted messaging.
Subway P2Pyou and your agents work normally. tenet captures everything — decisions, outcomes, patterns. over time, it builds a world model of your project and starts improving it autonomously.
you and your agents work normally using any tool. tenet quietly captures every decision, every code change outcome, every pattern. journals accumulate. memory indexes. context hub serves it all via MCP.
tenet knows your naming patterns, your architecture preferences, which approaches work in YOUR codebase. agents get better suggestions from memory search. you notice: "it remembered that decision from last week."
the policy head has enough training data. RL agents try improvements while you sleep — eval against your metrics, keep what works, revert what doesn't. you wake up to pull requests that actually make sense for your project.
the world model deeply understands your project — not just what the code does, but how it was built and why decisions were made. new team members' agents inherit the full context from day one.
one tenet workspace. your agents learn your patterns, your preferences, your codebase. the policy head trains on YOUR decisions. overnight agents improve YOUR metrics.
tenet init
parent tenet workspace scopes child workspaces per service. each service has its own context, agents, and eval. parent sees aggregated health. new hire's agents inherit full context from day one.
tenet init --parent ./platform