A2CR
A2CR solves the AI handoff problem: when your context window fills up or you switch between Codex, Claude Code, or Roo Code, you lose the thread. This server gives you save_context and resume_context tools that store compact WorkBaton checkpoints (goal, current state, next action, decisions, blockers) and optional WorkStash notes. Everything is client-encrypted before hitting the hosted A2CR service at a2cr.app, so you need an API key. The Python wrapper ships via PyPI as a2cr-mcp, and there's an MCPB package for Claude Desktop. Use it when a coding task spans multiple sessions and you want structured resume state instead of pasting transcript snippets into the next window.
- GitHub stars
- 2
- tools
- 13
- transport
- STDIO
- env var
- 1
- last updated
- Jul 1, 2026
Tools
13 toolsPublic tool metadata for what this MCP can expose to an agent.
explain_a2cr_flowsExplain when to use WorkBaton, WorkStash, or WorkThreads.
Explain when to use WorkBaton, WorkStash, or WorkThreads.
No parameter schema in public metadata yet.
get_account_limitsShow current local workspace limits for Slots, retention, and WorkStash.
Show current local workspace limits for Slots, retention, and WorkStash.
No parameter schema in public metadata yet.
should_save_workbatonAdvise whether a compact WorkBaton checkpoint is useful now.
Advise whether a compact WorkBaton checkpoint is useful now.
No parameter schema in public metadata yet.
save_contextSave a WorkBaton checkpoint in the local workspace.
Save a WorkBaton checkpoint in the local workspace.
No parameter schema in public metadata yet.
resume_contextFind and load the right WorkBaton for a fresh AI window.
Find and load the right WorkBaton for a fresh AI window.
No parameter schema in public metadata yet.
load_contextLoad a specific Slot number or named WorkBaton.
Load a specific Slot number or named WorkBaton.
No parameter schema in public metadata yet.
list_contextsList active WorkBaton Slots.
List active WorkBaton Slots.
No parameter schema in public metadata yet.
delete_contextDelete a named WorkBaton Slot.
Delete a named WorkBaton Slot.
No parameter schema in public metadata yet.
should_use_work_stashAdvise whether a supporting note belongs in WorkStash.
Advise whether a supporting note belongs in WorkStash.
No parameter schema in public metadata yet.
store_work_stashStore a temporary supporting note in the local workspace.
Store a temporary supporting note in the local workspace.
No parameter schema in public metadata yet.
get_work_stashRetrieve a referenced WorkStash entry.
Retrieve a referenced WorkStash entry.
No parameter schema in public metadata yet.
list_work_stashList WorkStash metadata and quota usage.
List WorkStash metadata and quota usage.
No parameter schema in public metadata yet.
delete_work_stashDelete a WorkStash entry that is no longer needed.
Delete a WorkStash entry that is no longer needed.
No parameter schema in public metadata yet.
Sponsors
Agent CaffeineSponsoredClose your lid, come back to finished work
SnugSponsoredThe clipboard manager for the AI eraPromote your product hereReach 125k+ visitors monthly
WaxSponsoredBlock anything with supervision, without erasing your iPhonePromote your product hereReach 125k+ visitors monthly
knobsSponsoredSee your site on any phone, any day, in any languageConfiguration
A2CR_LOCAL_DBOptional path for the local A2CR SQLite database. Leave unset to use the OS default app-data location.
Readme
A2CR logo
A2CR
A2CR is an open-source local MCP workspace for AI agent handoffs. It lets Codex, Claude Code, Cursor, and other MCP-capable agents save WorkBaton checkpoints, store temporary WorkStash notes, coordinate through WorkThreads, and resume long coding work from a fresh AI window.
Long AI work usually breaks at the handoff. A fresh AI window needs the goal, current state, decisions, blockers, validation, and next action, but not a whole noisy transcript. A2CR keeps that handoff state compact, explicit, and safer to share between sessions.
Use A2CR when you want to:
Details
- Repository
- github.com/a2cr/a2cr
- Category
- AI & LLM Tools
- Package
a2cr-mcp- Transport
- STDIO
- Registry
- MCP registry, active
Related AI & LLM Tools MCP servers
View all
second-brain MCPddmanyes/second-brain-mcpSelf-maintaining knowledge vault: figure-level search, auto-wikilinks, and memory compression.2stars
Youtube Video Analyzerludmila-omlopes/youtube-video-analyzer-mcpMCP stdio server for analyzing YouTube videos with Google Gemini2stars
Sentinal Mcpmusaraf-m/sentinalRedis health, BullMQ queue monitoring, memory analysis, and slow query diagnostics2stars
Kernal Mcppintomatic/kernalGraph-native memory for AI agents: a knowledge graph built from conversation, via MCP.2stars
Sprintrasprintra-io/sprintra-mcpThe project brain for AI coding agents — memory, decisions, sprints, knowledge base via MCP.2stars
fmrltoogreatwtf/fmrl-pluginYour agent's outbox. Ask it to publish to fmrl and get a shareable link, no account required.0stars