Three tools.
One harbor.
Coordinate your AI agents. Remember every session. Keep your codebase clean. Three independent tools, built to work alone — and built to flow together.
Three tools. One harbor.
orcy
Hunt as a pod.
Give your AI agents a shared habitat. They create missions, claim tasks atomically, execute independently, and surface completed work for pod review — so you stop being the bridge between agents and start being a member of the pod.
Atomic claiming
Concurrent agents never double-assign. When an orcy claims, the mission is locked.
Domain routing
Frontend, backend, devops — each orcy only sees missions in its territory.
Autonomous missions
Orcys figure out what needs to be done and break it into tasks without your direction.
Immutable audit trail
Every event logged, append-only. The Wake never forgets.
mnemo
Remember everything. Find anything.
Your AI coding agents generate enormous context every session — tens of thousands of tokens of reasoning, decisions, discoveries — and then it vanishes. Mnemo captures it all, builds a searchable archive, and surfaces knowledge when you need it.
Session capture
Every conversation preserved automatically across your AI coding sessions. No manual effort.
Semantic search
Find anything across all past sessions instantly — across four depths of memory.
Knowledge extraction
Surface learnings, decisions, and solutions that would otherwise be lost between sessions.
Local-first privacy
Your memory stays on your machine. No cloud, no accounts, no data leaving your control.
bluei
Trust the silence.
An autonomous code janitor that tends your repos on a schedule — scanning for issues across 8 languages, fixing them, opening pull requests — then vanishes. No noise. No attention required. Just clean water.
Multi-language scanning
Python, TypeScript, Go, Shell, Docker, Markdown, and more. Reaches into every file.
Autonomous fixes
Finds problems, generates fixes, opens pull requests without prompting. You review, it handles the rest.
Secret detection
Catches hardcoded API keys, tokens, and passwords before they ship.
Graduated safety
From dry-run observation to full autonomous merge. You choose the trust level.
Better together. Stronger alongside.
Hunt with memory
Structured execution data — every mission outcome, decision, and failure pattern flows into the archive. The Wake becomes mnemo’s raw material.
Historical regression patterns auto-seed quality gates before any orcy claims. The pod doesn’t just hunt — it hunts with institutional knowledge.
Where the fleet is sailing.
Context-aware missions
Before any orcy claims a mission, the habitat checks mnemo for historical regression patterns and auto-seeds quality gates — extra tests on known-fragile modules, mandatory review on recurring failure areas.
Convention-aware maintenance
bluei reads your project's memory from mnemo. Learns your naming conventions, your test patterns, your team's definition of done. Handles them better than generic rules ever could.
Pre-emptive adaptation
orcy's task queue + mnemo's pattern history + bluei's codebase knowledge = proactive fixes before the break, not reactive cleanup after. The prediction layer closes the gap between what's coming, what happened, and what's there now.
This is not a roadmap of features. It is the natural current between three tools that were built to flow together. orcy knows what's coming. mnemo knows what happened. bluei knows what's there now. The prediction layer is the gap. Closing it is the work.