The hivemind for human–agent teams. SageOx makes your team's decisions, conventions, and architectural intent persistent — and loads them automatically into every coding session, human or AI.
Today, that context lives in scattered places: a meeting nobody recorded, a Slack
thread three weeks deep, the head of whoever wrote the code. So humans repeat
themselves and AI coworkers drift — rebuilding the same lost context every
session. ox closes that gap. Decisions become shared memory; every session
starts with the full picture instead of from zero.
Ask your coding agent what the team already figured out — even if it happened in a different agent, on a different machine, days ago.
One session, every moment that matters: it recalls a teammate's Codex work and your own prior Claude Code session, coordinates by murmuring and noticing a teammate already in the same files, and proposes a plan enriched with collisions, prior art, and who owns the area. All of it because every session is recorded to a shared, queryable Ledger.
flowchart LR
REC["Record at sageox.ai<br/>discussions, decisions"]
KB["Knowledge Bubbles<br/>team context plus ledger"]
PRIME["ox agent prime<br/>loads context"]
WORK["AI Coworker session"]
CAP["Session captured<br/>to the Ledger"]
REC --> KB --> PRIME --> WORK --> CAP --> KB
Record a discussion once. It's distilled into Knowledge Bubbles — your team's shared memory. The next time anyone primes a coding session, that context flows in. The session itself is captured back to the Ledger, so the next coworker inherits it too. Context compounds instead of evaporating.
ox is agent-agnostic. The same recorded context is primed into, and queryable
from, whichever agent your team uses.
| Agent | Prime context | Record sessions | What fires it |
|---|---|---|---|
| Claude Code | ✅ | ✅ | hooks — session, prompt, tool, and compaction |
| Codex CLI | ✅ | ✅ | hooks — session, prompt, and tool |
| Gemini CLI | ✅ | ✅ | hooks — tool and stop |
| Droid | ✅ | ✅ | hooks — tool and stop |
| OpenCode | ✅ | ✅ | plugin — session start |
| Amp | ✅ | ✅ | plugin records; AGENTS.md primes |
| Pi | ✅ | ✅ | AGENTS.md |
| Aider | ✅ | ✅ | CONVENTIONS.md |
| Goose | ✅ | ✅ | hooks — session, prompt, tool, and tool-failure |
| Cursor · Windsurf · Cline · Copilot · Kiro | ✅ | ➖ | instruction file |
✅ shipped · ➖ planned. Claude Code is the primary, most-tested target; see agent compatibility for the per-agent detail.
ox query, ox murmur, and ox status are plain CLI commands. They work from
any agent — including ones with no adapter — as long as ox is on PATH.
ox also detects the harness running your agents, and adapts what it tells them:
- Block Buzz — spawns agents through its
buzz-acpACP harness - Conductor · Gas City · OpenClaw
Detection is config-independent — ox walks the process ancestry rather than
trusting an environment variable, so it works even when the harness sets nothing.
Under Buzz, ox agent prime emits an extra directive: buzz-acp does not fire
the agent lifecycle hooks that push whispers into a turn, so the agent is told to
pull teammates' signals by hand instead of silently missing them.
Agent and orchestrator names are trademarks of their respective owners; ox
is compatible with, not affiliated with, them.
Quick install (macOS / Linux / FreeBSD):
curl -sSL https://raw.githubusercontent.com/sageox/ox/main/scripts/install.sh | bashFrom source:
git clone https://github.com/sageox/ox.git && cd ox
make build && make installVerify with ox version.
cd ~/src/my-project # your code repo
ox login # authenticate with sageox.ai
ox init # one-time per repo: creates .sageox/
git add .sageox/ && git commit -m "initialize SageOx"
ox doctor # diagnose (ox doctor --fix auto-repairs)
ox status # setup, sync state, and your Knowledge BubblesThat's the whole setup. From here, session capture is automatic — when an AI
coworker runs ox agent prime at the start of a session, team context loads in
and the session is recorded to the Ledger when it ends. No manual start/stop
ritual.
Then just ask your AI coworker things it couldn't have known on its own:
- "What did we decide about the daemon design, and who worked on it?"
- "Draft a plan from this week's SageOx team discussions."
- "Show me an effective prompt a teammate used on this repo recently."
| Capability | Command | Learn more |
|---|---|---|
| Team context + repo memory as Knowledge Bubbles | ox status, ox kb list |
jit-discovery |
| Query past sessions, discussions, and code — across agents and machines | ox query "..." |
query |
| Auto-recorded coworker sessions (Ledger) | ox agent prime, ox session list |
session capture |
| Real-time coordination signals between coworkers | ox murmur "..." |
— |
| Team-context-enriched implementation plans | ox plan enrich, ox plan render |
plan |
| Planning-relevant code insights (hotspots, contention) | ox code insights |
— |
| Load an expert AI coworker into context | ox coworker load <name> |
— |
| Diagnose and auto-fix your setup | ox doctor --fix |
doctor |
ox init writes a .sageox/ directory that ties the repo to your team. From
then on, ox agent prime injects team context — conventions, security
requirements, architectural decisions, prior sessions — into each AI coworker
before it writes a line. Coworkers, human and AI, share that context through the
Team Context and the per-repo Ledger.
The payoff is multiplayer by default. When a discussion, an implementation session, and the resulting code all carry the same shared context, a reviewer opening the PR has the full story — the original reasoning, the session that built it, and the diff — without chasing anyone down.
SageOx reads configuration from, in order:
- CLI flags (
--verbose,--quiet,--json) - Environment variables
- Config file (
.sageox/config.yaml)
We build ox in great company. These are the tools we rely on — and love —
every day. Gratitude to the teams behind them, and to the wider developer
community.
Socket · SageOx · CodeRabbit · Greptile · Charm

