Sync — Take AI agents under absolute control · Sync
Free & private — star it on GitHubTake AI agents under absolute control<br>Sync is the project-management and quality-control layer your AI agents work against. Define your specs, rules, decisions, and roadmap once — Sync holds every agent accountable to them, so they build against your architecture instead of hallucinating or repeating past mistakes. When the code moves, stale context gets flagged and the agent has to stop and ask.<br>Download
View on GitHub<br>From install to results in one step. Runs entirely on your machine — nothing is ever uploaded. Currently in alpha · v0.4.5. Building team & cloud features — join the waitlist.<br>MCP server · Desktop app · Built-in terminal · Claude Code · Cursor · Zed · Codex · OpenCode
sync — workspace
Bad context is how you and your agent drift apart<br>An agent is only as aligned as what it reads. Point it at stale, scattered context on a large task and it heads off course — confidently, while you're looking the other way.
Pattern-matches stale text<br>It trusts whatever it finds, even when the code moved on months ago.
Hallucinates the gaps<br>Where docs are missing, it invents plausible-but-wrong context.
Re-introduces dead decisions<br>It re-litigates choices you already overturned, again and again.
Get started in seconds<br>Three steps to agent control<br>No mandatory terminal commands, no git clone, no long configuration files. Everything happens in the desktop app or with a single click.
01<br>Install the desktop app<br>Download Sync for macOS, Windows, or Linux. One binary, no dependencies — just open it.
02<br>Open your project folder<br>Point Sync at any local repository. It sets up the .sync/ knowledge base automatically — no manual config.
03<br>Connect your favorite agent<br>The Welcome wizard wires up Claude Code, Cursor, Zed, Codex, or OpenCode instantly — and you can run them right inside Sync's built-in terminal. No config files.
Download
Latest release v0.4.5 · macOS, Windows & Linux · free and open source
WorkspaceApi<br>Search…⌘K
Decisionvalid<br>Use queue-based ingestion for uploads<br>Decouples the API from processing so a slow worker never blocks a request. Supersedes the inline approach.<br>src/server/ingest.tslinked to 2 constraints
Constraintstale<br>Uploads must stay under 10 MB<br>The file it describes changed in the last commit — Sync flags it for review before your agent trusts it.<br>src/server/upload.tstouched 2h ago
sync · alpha · a1b2c3dM1 read-only · v1
How Sync is different<br>Built to govern agents, not just feed them text<br>AI coding agents are powerful, but without a plan to work against they cause architectural drift, forgotten decisions, and regressions. Memory files and wikis can hold context — they aren't built to keep an agent accountable to your project. Sync is: a mission-control layer above your editor and agent.
Sync
Project management & quality control<br>Memory systems
e.g. Cursor rules<br>Documentation
Confluence · Notion<br>Primary goalControl agent results & enforce quality<br>Provide simple text context to the agent<br>Human-readable knowledge sharing<br>StructureStrict, agent-readable constraints & widgets<br>Free-form text files<br>Unstructured text, tables, images<br>Staleness trackingAutomatic — tied tightly to code changes<br>None / manual<br>Manual audits required<br>Agent behaviorMust respect the rules; asks for approval when context goes stale<br>Reads notes, but often ignores them<br>Usually can't access it seamlessly<br>User experienceVisual Project Hub, widgets, direct oversight<br>Simple text editing<br>External web interface
Sync works as a highly effective memory system too — but its real power is governance and quality control.
Why Sync<br>One source of truth you and your agent share<br>Structured, versioned, and always current — so you and your agent (and the whole team) build the same thing, instead of you cleaning up after it drifts.
Automated quality gates<br>Knowledge that flags itself stale<br>Every entry is bound to the code it describes. When a commit moves past it, Sync marks it stale — pointing at the exact commit and files — and makes the agent stop and ask for your input instead of silently building on old data.<br>Bound to the code; flagged stale on the commit that moves past it<br>valid · unverified · stale · invalid — never a guess<br>Forces the agent to pause and ask before trusting stale context
.sync/decisions/d-3c2a4f.md<br>decision · d-3c2a4fvalid<br>Sessions are signed JWTs with a 15-minute expiry.<br>scope · src/auth/session.ts
Written as it works<br>Knowledge your agent writes<br>Specs, decisions, constraints, and observations get captured during coding, by the agent, in typed structure it can query precisely later — not prose nobody re-reads.<br>Decisions with the reasoning attached<br>Constraints the agent must respect<br>Observations about how the system really behaves
agent · git-sync mcp<br>›Reading src/auth/* before changing it…<br>→ writing decision: token strategy<br>→ writing constraint: no PII in logs
knowledge written<br>decision · d-7af213valid<br>Use...