SenrelayProvider Presets & Local Config Sync
Switch target hosts from the top bar, fill built-in presets in one click, run live endpoint latency tests, and preview-import existing local configurations on demand.
CONTEXT IS CONTINUITY
Switch agents,
keep the thread.
Switch tools.
Keep your thinking.
Senrelay is a local multi-agent context workbench for macOS 14.0+ (Apple Silicon) developers—100% free and open source ($0, MIT License, no account required).
Type /contextcore or say "Relay" across Claude Code, Codex, Cursor, Antigravity, OMP, and OpenCode to scan recent conversations and continue seamlessly; capture source-linked memories as you chat (with a brief 💡 receipt); and manage providers, MCP, and Skills in one place.
Multi-host configuration management·Implicit auto-memory·Cross-agent relay·MIT open source
FIG. 01 SHARED CONTEXT FIELD
SENRELAY
CONTEXT IN MOTION. Illustration · not live statusCONTEXT / 001
01 / THE REASONKeep context with the project
Spend time moving forward,
not re-explaining.
Architecture settled, code halfway done.
New agent, another briefing from scratch?
Senrelay brings 6-point cross-agent conversation relay, source-linked companion auto-memory, and multi-host MCP/Skills configuration management into one local workbench on your Mac. During chats, companion MCP tools and Skills capture structured memories with source links (leaving a brief 💡 receipt); switch to any new assistant and type /contextcore or say "Relay" to scan recent conversations, pick a session, or pass a follow-up task to continue right away (see the core workflows guide).
[ 01 ]
Provider and model switching
Switch hosts from the top bar, fill provider presets in one click, test latency, switch or sync MCP and Skills across hosts, and import existing local configurations on demand.
[ 02 ]
Companion Implicit Auto-Memory
Companion MCP tools and Skills capture structured memories with exact source links and a brief 💡 receipt during chats, alongside manual editing and multi-format document imports.
[ 03 ]
One-phrase cross-agent relay
Type /contextcore or say "Relay" in a new agent to list recent conversation candidates; reply with just a number for a brief summary report, or include a task to continue immediately.
Built for everyday multi-agent workflows
- Cross-Host Coding RelayFinish backend APIs in Claude Code, then switch to Cursor or Codex with exact diff boundaries and test state.
- Research & ArchitectureCarry multi-turn sources, ruled-out options, and constraints into another model to keep reasoning.
- Implicit Auto-MemoryCapture coding conventions, tone preferences, and lessons learned with source links and a brief 💡 receipt during chats.
- Unified Configuration ControlSwitch and sync providers, MCP, Skills, and system prompts in one click with automatic rollback snapshots and external-edit conflict diff prompts.
02 / AGENT MANAGEMENTModels, tools, and preferences in their own place
Your agents.
Unified here.
One-stop provider, model, tool, and prompt management across your agents. Switch Claude Code, Codex, Cursor, Antigravity, OMP, and OpenCode—plus extended hosts—from the permanent top bar (see the 6+9 host support matrix & pricing fact sheet), run one-click Senrelay onboarding, and switch or sync resources with automatic rollback snapshots and external conflict diff review (local data security & Keychain isolation).
- 01 / SWITCH HOSTSelect target host in top bar
- 02 / PRESET & SPEEDTESTApply presets & test latency
- 03 / SYNC RESOURCESSync MCP, Skills & instructions
- 04 / GUARD & APPLYConflict check & safe rollback
Multi-Host ConsoleClaude CodeCodexCursorAntigravityOMPOpenCode
Local Config Sync · Diff Preview
One-Click Switch & PresetsOfficial / OpenRouter / DeepSeekBuilt-in provider presets · 142ms illustrative latency · Keychain isolation
Cross-Host Resource SyncMCP Tools · Skills · Global PromptsConfigure once and sync to selected Agent hosts in one click
Conflict Guard & RecoveryWrite Verification + Env AlertsDetects shell environment overrides with snapshot diff and rollback
Multi-Mode Routing & Keychain
API keys stay isolated in macOS Keychain. Switch each assistant between Direct, Local Proxy Routing, or Model Aggregation with ordered automatic fallback chains.
MCP Tools & Companion Skills
Configure stdio or HTTP tool servers and import Skills from local folders, ZIPs, or GitHub. Ships with companion auto-memory and cross-agent relay components with host load verification.
System Prompts & Project Bundles
Maintain coding conventions and global rules in one place, or bundle project-specific providers, models, MCP tools, Skills, and prompts to preview diffs and switch in one click.
Request Traces, Usage & Latency
Track request counts, token consumption, first-token latency, and cost estimates in real time, clearly separating primary calls from retry chains across long-term trends.
External Edit Guard & Safe Snapshots
Clearly distinguishes applied state, pending updates, and external manual edits. Every write saves a recovery snapshot with diff inspection, conflict alerts, and instant rollback.
Each host uses its own native configuration format and capability scope; the workbench provides live write verification, latency checks, and environment conflict alerts to keep every change predictable.
See assistant support03 / CROSS-AGENT PROJECT RELAYContext in place, thinking in place
Switch agents.
Keep the project context.
Type /contextcore or say "Relay" in any new agent to list recent conversation candidates across hosts (with number, source agent, project, title, and summary); pick just a number to review a brief summary and wait for your next prompt, or attach a specific task to continue right away.
What the next assistant needs to know.
Scan recent cross-host conversations → review the numbered candidate list → get a status summary or continue immediately with these six handoff elements:
STEP 01 · TRIGGER RELAYType /contextcore or "Relay" in your new agentAutomatically scans recent conversations across connected hosts without manual copy-paste
STEP 02 · PICK CANDIDATEView index, source agent, project, title & summaryLists recent cross-host conversation candidates chronologically with clear context
STEP 03 · REPORT OR CONTINUEPick a number to brief & wait, or add a task to startLoads the relay pack and recent transcript turns to match your workflow pace
Project and task goal
What needs to be done, and what is outside the scope.
Current progress
What is finished, and what still needs work.
Decisions and constraints
Agreed decisions and boundaries to preserve.
Working scene and evidence
Original conversations, project records, and reference materials.
Blockers and open questions
Open problems and assumptions that need checking.
Next steps
Where to resume and how to check the result.
04 / AUTO-MEMORY & WORKSPACEBelongs to the project, not one chat
The story of the project,
always within reach.
01 — PROJECTScattered chats,
Scattered chats,
back in one project.
Stop digging through scattered chat windows. Sessions, progress, decisions, and references from every agent converge in one project with direct source links.
Organized by project · search across sourcesPaper BridgePROJECT WORKSPACE
PreviewOverviewChatsMemoryMaterials
Keep the existing design system,
so iterations stay consistent.
Background ↗Tasks ↗Lessons ↗
02 — AUTO-MEMORYLet experience accumulate,
Let experience accumulate,
not chats pile up.
No manual note-taking required—companion MCP tools and Skills capture preferences, background, decisions, progress, and lessons with source links and a brief 💡 receipt during chats, with 📌 core-rule pinning and conflict review.
Companion MCP/Skills · 💡 receipt · Source-traceablePreferencesBackgroundConstraintsTasksLessonsReferences
03 — KNOWLEDGEKeep your documents
Keep your documents
working for you.
Import requirements, guidelines, and references for authorized assistants to search. Review the source and extracted text in the workspace.
Markdown · TXT · JSON · PDF · DOCX · DOCPROJECT KNOWLEDGEEvery document,
has a next chapter.
.pdf.md.docxhas a next chapter.
05 / START WITH ONE TASKOne assistant, one task to start
Ready in four steps.
Start relaying.
Launch the local service and workbench to unify multi-host configurations and companion components; capture source-linked project memories during daily chats and relay across agents with a single command.
- 01
Start Local Workbench
Open the desktop app or follow the local setup guide to start the local service, and optionally preview and sync existing local configurations.
- 02
One-Click Connect & Sync
Switch hosts in Agent Management, click One-Click Connect to wire history paths, MCP, and /contextcore Skills, and enable providers with live latency checks.
- 03
Capture Auto-Memories
Work normally in your favorite agent while companion MCP tools and Skills save source-linked structured memories with a brief 💡 receipt.
- 04
Type "Relay" in New Agent
In a new agent, type /contextcore or say "Relay" to pick a recent conversation candidate for a brief status report, or attach a follow-up task to continue immediately.
Local first, clear permissions.Understand data and permission boundaries+
All session indexes, project memories, and config snapshots stay on your Mac, with API keys isolated in Keychain. Reading chat history, editing host configs, and downloading optional semantic models are confirmed separately, and search never connects to the network.
06 / IN YOUR OWN TOOLSKeep your way of working
Your familiar agents.
Connected to the same project.
Keep using familiar tools. Senrelay stores and provides shared project context.
- CodexDesktop / CLICarry progress into the next stretch of work
- CursorIDEPick up existing thinking in the editor
- OMPOh My PiLet context follow your workflow
- AntigravityDesktopChange environments, keep the project background
- Claude CodeCLIContinue the previous stretch in the terminal
- OpenCodeCLIBring project context into terminal workflows
The six core agents above support full session capture, implicit auto-memory, and cross-agent relay; additional extended hosts support configuration file management only.
| Tier & Runtime | Covered Hosts / Components | Supported Capabilities & Technical Specs |
|---|---|---|
| 6 Core Context Hosts | Claude Code (CLI), Codex (Desktop / CLI), Cursor (IDE / CLI), Antigravity (Desktop), OMP (Oh My Pi CLI), OpenCode (CLI) | Full local session history capture, companion MCP/Skills implicit auto-memory (with brief 💡 receipt), /contextcore cross-host conversation relay (number-only selection sets must_not_execute=true to brief and wait), AGENTS.md handbook preview, and host config sync (e.g. ~/.claude.json, ~/.codex/config.toml). |
| 9 Extended Config Hosts | Claude Desktop, Gemini CLI, Grok Build, Kimi Code, MiniMax Code, Pi, Hermes, OpenClaw, agy | Previewed configuration file management (JSON / JSON5 / TOML / YAML), diff inspection, rollback snapshots, and writes only. No session history capture, companion auto-memory, or cross-host relay. |
| Local Runtime & Storage | macOS 14.0+ · Apple Silicon (arm64) | Strictly bound to loopback 127.0.0.1:8420 (dev mode 127.0.0.1:18420); local SQLite + FTS5 full-text search (0600 DB file mode) with API keys isolated in macOS Keychain. |
| Dimension | Official Fact & Plain-Text Pricing | Verification Details & Real Boundaries |
|---|---|---|
| Pricing & License | $0 / 100% Free & Open Source (MIT License) · No Account Required | Permanently free ($0) with no subscription tiers, no cloud paywall, and no email or account signup; you bring your own model provider API keys stored locally in macOS Keychain. |
| Current Version & Install | v1.0.0-beta.1 · brew install --cask qingsongcui/tap/senrelay | Install via Homebrew Cask, download Senrelay-macOS.zip from GitHub Releases, or build from source (see first-launch steps). Do not run pip install contextcore (unrelated third-party PyPI package). |
| Known Limitations & When Not to Use | macOS 14.0+ arm64 only · Local single-Mac only · Beta needs xattr -cr | When NOT to use Senrelay: (1) You use Windows, Linux, or an Intel (x86_64) Mac; (2) You need multi-device cloud sync or hosted team SaaS collaboration; (3) Public Beta builds use ad-hoc signing without Apple Developer ID notarization, requiring a one-time xattr -cr /Applications/Senrelay.app on first launch (see our 5-way multi-agent context & MCP comparison). |
Multi-host integration & capability coordination
The six core hosts—Codex, Claude Code, Cursor, Antigravity, OMP, and OpenCode—support full local session capture, companion implicit auto-memory, and cross-agent conversation relay. Additional extended hosts support previewed configuration file management only.
Configuration file writes, environment variable conflict checks, endpoint latency tests, and live process reload verification are tracked independently so status stays transparent.
07 / A FEW GOOD QUESTIONS
A few more questions.
About handoffs, data,
and what you can do today.
How does Cross-Agent Conversation Relay work in practice?
Type /contextcore or say "Relay" in your new agent to list numbered recent cross-host conversations and resume by index. When switching between Claude Code, Cursor, Codex, Antigravity, OMP, or OpenCode, the companion Skill and local service show each candidate's index, source agent, project, title, and summary: replying with just a number loads the relay context (must_not_execute=true) to report a status summary and wait; attaching a specific task continues working immediately.
How is implicit auto-memory triggered automatically?
Companion MCP tools and Skills automatically capture structured memories with source links and a brief 💡 receipt during everyday chats—no manual note-saving required. When discussions produce architectural decisions, user preferences, constraints, task milestones, or lessons learned, they are recorded with verbatim turn provenance; conflicting updates route to the Pending Conflicts inbox, and you can edit memories, pin 📌 core rules, or import documents anytime.
Which local configurations can I sync or import?
You can switch and sync providers, models, MCP servers, Skills, and prompts across 6 core hosts and 9 extended config hosts, or import existing local config files. Senrelay includes top-bar host switching, built-in provider presets, one-click activation with automatic rollback snapshots, endpoint latency testing, model mapping, and shell environment conflict detection.
Will changing one assistant overwrite my other assistants?
No—changing one assistant never silently overwrites another assistant's configuration. You can switch providers or rules for a single host, or explicitly check multiple hosts to sync shared MCP tools, Skills, and prompts with full diff preview and automatic rollback snapshots.
Do my transcripts and API keys stay safe on my Mac?
Yes—100% of your transcripts, memories, and config snapshots remain on your local Mac and are never uploaded to the cloud. All data lives in your local SQLite database (0600 file permissions), and API keys are isolated in macOS Keychain and excluded from config exports by default.
How much does Senrelay cost? Do I need an account or subscription? (Pricing)
Senrelay is 100% free ($0) under the open-source MIT License, with no account registration, no subscription tiers, and no cloud paywall. Every cross-agent relay, auto-memory, and multi-host configuration feature runs locally without feature gates; you only provide your own model provider API keys in your AI coding assistants.
What are Senrelay's real limitations, and when should I NOT use it?
Do not use Senrelay if you work on Windows, Linux, or Intel Macs, or if you require multi-device cloud sync and hosted team collaboration. Current release v1.0.0-beta.1 runs exclusively on single-machine macOS 14.0+ (Apple Silicon arm64); the 9 extended hosts support configuration file management only (not session relay); and Public Beta builds require a one-time xattr -cr /Applications/Senrelay.app command on first launch before Apple notarization is added. If you are weighing alternatives, try our 5-option comparison & 30-second stack picker.
How do I install and launch the workbench right now?
Run brew install --cask qingsongcui/tap/senrelay via Homebrew or download the free MIT-licensed Senrelay-macOS.zip from GitHub Releases. You can also follow the local setup guide to run from source on loopback port 127.0.0.1:8420 / 18420. Note: never run pip install contextcore—the PyPI package of that name is an unrelated third-party project.
Next time,
start from here.
Launch your local multi-agent workbench in minutes.
Manage multi-host configurations, capture source-linked memories as you chat, and relay conversations across agents anytime.
Homebrew Cask:
brew install --cask qingsongcui/tap/senrelayAbout the Creator & Open-Source Project
Senrelay (open-source repository ContextCore) is created and maintained by independent developer Qingsong Cui (George O'Nair) (GitHub @qingsongcui · X @George_ONair) to solve context loss and configuration drift when switching between Claude Code, Codex, Cursor, and other AI coding agents.
Local Privacy & Legal Notice
Senrelay runs strictly on local loopback 127.0.0.1:8420 with zero cloud telemetry, zero transcript or code uploads, and no account registration; all SQLite databases (0600) and macOS Keychain credentials stay on your Mac. Distributed "AS IS" free of charge under the MIT License.
macOS 14.0+ (Apple Silicon) local service + workbench · v1.0.0-beta.1 · $0 MIT free & open source · Fact-checked & updated .