Keep local AI agents from starting over.
Memcore Cloud is not another summary-to-vector memory API. It is a local AI memory library that keeps original records, library ids, Zhiyi preferences, Xingce work paths, Hermes skill-to-experience candidates, and borrowing receipts on your own machine.
Memcore Cloud is the English product name. 忆凡尘 / Yifanchen remains the Chinese name and codename.
Most memory layers stop at add and search. Memcore Cloud goes further: AI work should keep original evidence, remembered items should have a shelf, and useful experience should be able to improve without losing its source.
Your AI conversations, project decisions, fixes, and corrections should not be scattered across disposable chat windows or collapsed into untraceable summaries. Memcore Cloud stores them as local, source-backed library holdings: each useful memory can keep a source trail, original text, a library id, and a receipt showing how it was used.
Xingce experience is not locked inside one agent. Any local platform that can use a skill, custom instruction, or MCP tool can receive the same work experience before it acts: what worked before, what failed before, which checks proved it, and whether the task should be repeated at all.
After connection, an agent can recall who you are, what you prefer, which project boundaries matter, what was decided before, which fixes worked, and which mistakes should not be repeated. When it uses memory, you can see what it borrowed before it answered.
This is not just chat-history search. It is a library system for local agents:
- raw records are the original holdings;
- library ids locate memories, preferences, tool facts, and work paths;
- Zhiyi stores identity clues, preferences, corrections, habits, and project boundaries so agents understand you better over time;
- Xingce stores proven repair paths, troubleshooting order, validation steps, gotchas, and work methods so agents get better at doing the work;
- common experience entry points let skill-, instruction-, or MCP-capable local agents read the same Xingce guidance before work;
- Hermes skill-to-experience diff compares skill files with Xingce experience and creates reviewable adoption or upgrade candidates;
- receipts show what the agent read before it acted.
Experience evolves, but it stays traceable. A successful fix, a mistake, or a user correction can enter the candidate shelf first; only source-backed items with original evidence and acceptance checks can be adopted into Xingce. Later changes can leave errata, upgrade receipts, or rollback receipts.
- Shared local context: use one local record base across Claude Desktop, Claude Code CLI, Codex, OpenClaw, Hermes, Cursor-style tools, and popular open-source agents.
- Raw records first: keep original conversations, tool output, source tool, device, and timeline before any summary; summaries help navigation but do not replace the holding.
- Library ids and borrowing receipts: useful memories can carry source trails, original text, collection identity, and a receipt showing what was used.
- Zhiyi understands you: preserve identity clues, preferences, wording habits, corrections, and project boundaries so the next window asks fewer repeated questions.
- Xingce improves work: preserve repair paths, troubleshooting order, validation steps, gotchas, and working methods so proven approaches can be reused.
- Experience for every local agent: deliver the same Xingce experience through skills, custom instructions, MCP, and
work_preflightso Codex, Claude, OpenClaw, Hermes, Cursor-style tools, and other local agents can check work history before acting. - Hermes skill evolution: compare Hermes skills with Xingce experience in a read-only diff, then turn new skills or changed skills into reviewable adoption or upgrade candidates.
- Source-backed recall: ask an agent about old decisions, preferences, fixes, or project boundaries and get compact source refs, library ids, hit reasons, and optional bounded excerpts.
- Pre-work context check: before coding, installing, syncing, or troubleshooting, let the agent check whether the answer is already in your local record base.
- Record Doctor: run a safe one-command check before recall to see whether source records, raw mirrors, the canonical index, and memory/experience links are guarded.
- Local console: open a browser page to see connected tools, recent record health, safe capability checks, and where new raw records are stored.
- No cloud account required: local data stays on your machine by default. Summaries help navigation, but original records remain the source of truth.
- Simple install options: use one shell command, PowerShell, or the double-click installers included in the release zip.
After install, open the local console:
http://127.0.0.1:9850
Then run the safe first check:
{"query":"capability check","mode":"capability_check"}A healthy first result says the connection is read-only, no real memory was recalled, and no raw excerpt was returned. After that, try a real question such as:
What did we decide last time about this project?
Memcore Cloud is designed to answer with source refs first, then expand to original evidence only when you ask.
Before asking an agent to change code, install, sync, or troubleshoot, ask it to check local context first. The expected behavior is simple: tell you whether the work looks already built, miswired, missing a diagnostic, or truly missing, then inspect the repo and tools before editing.
AI tools forget the small things that make work smooth: your preferred wording, project boundaries, old mistakes, useful fixes, and where a task left off. Memcore Cloud keeps that trail on your own machine so a new agent window does not have to start from zero.
It is not a hosted chat app and not a summary vault. It keeps source records, source refs, corrections, and work experience together so memory can point back to the original words.
Experience evolves, but it is not a black box. Memcore Cloud supports evidence-backed curation with validation and receipts. Experience moves like a library curation workflow:
raw record
-> experience candidate
-> review queue
-> source and acceptance-check validation
-> authorized adoption or rejection
-> rollback, supersede, or upgrade receipt
That means a useful repair path can become reusable Xingce experience, while a bad or unsupported lesson can stay in review, move to errata, or be rolled back. The current system supports curated evolution; it does not claim fully autonomous self-training.
- Shared local context for your AI tools: Claude Desktop, Claude Code CLI, Codex, OpenClaw, Hermes, Cursor-style tools, and fast-moving open-source agents can connect to the same local record base.
- Working methods that survive the next window: preferences stay available, while proven ways of working become reusable guidance.
- Preferences and experience stay distinct: Zhiyi keeps preferences, corrections, and boundaries; Xingce keeps repair paths, validation steps, and work methods.
- Experience can intervene across platforms: Xingce is not private to one tool. Any local agent with a skill, custom instruction, or MCP entry can read the same experience candidates, gotchas, and acceptance checks before work.
- Fewer repeated fixes: before starting work, an agent can check whether you already built the feature, tested the path, or found the same wiring problem earlier.
- Receipts, not vibes: remembered items keep source trails so you can check why something came back.
- Original records stay original: summaries can help navigation, but they do not replace saved source text.
- Tool boundaries stay visible: Claude Desktop and Claude Code CLI are treated separately; relay or official-login records are not silently mixed.
- A safer first test: capability check proves the local read-only path works without recalling real memory or returning raw excerpts.
- Install once, then it finds your tools: Memcore Cloud discovers local AI tools, including popular open-source agents such as OpenCode, Goose, Aider, and OpenHands, connects usable local AI tool entries, and keeps new memory sources in a computer-first archive.
- Smarter recognition for new tools: if you have a model configured, or already have a local AI model setting Memcore Cloud can reuse, it can identify unfamiliar tools from metadata only; if not, local rules still work.
If you use Codex, Claude Code CLI, OpenClaw, Hermes, or another local agent that can run terminal commands, paste this prompt into it:
You are installing Memcore Cloud for me on this machine.
Repository: https://github.com/strmforge/memcore-cloud
Install and start Memcore Cloud. Then install Memcore Cloud Zhiyi as a standing memory rule for this agent, not just a one-time setup note. If this platform supports skills or custom instructions, add the Memcore Cloud Zhiyi skill/instruction. If this platform supports MCP, register the MCP tool named yifanchen-zhiyi at http://127.0.0.1:9851/mcp.
After setup, run only a safe capability check with zhiyi_recall:
{"query":"capability check","mode":"capability_check"}
Do not recall my real memory yet; do not recall my real memory until I ask.
Keep this rule active from now on. When my question depends on prior context, call zhiyi_recall before answering and use source refs by default; ask for raw excerpts only when I explicitly need original evidence text. Triggers include previous decisions, corrections, project boundaries, forgotten context, install/test/release status, and short follow-ups in ongoing work such as "next step", "what else", or "then what". Use active recall by default: current window/session first, then same project/workspace, same workstream/task, then stable preferences/tool facts. Use raw-pool/global only when I explicitly ask for that wider view. If explicit window-only recall says the current window/session is not bound yet, say that plainly; do not claim there is no memory. If the skill is installed but zhiyi_recall is not available, tell me the MCP/tool connection is missing instead of guessing from memory.
Before coding, installing, syncing, release-prep, or remote troubleshooting, run a read-only pre-work check:
{"query":"<the work to do>","mode":"work_preflight"}
Use it to decide whether the work looks already built but forgotten, already built but miswired, missing diagnostics, or actually missing. Treat that result as a starting point; still inspect the repo, tests, tools, and docs before editing.
Chinese prompt:
你正在帮我在这台机器安装 Memcore Cloud（忆凡尘）。
仓库：https://github.com/strmforge/memcore-cloud
请安装并启动 Memcore Cloud。然后把 Memcore Cloud Zhiyi 安装成这个 agent 的长期记忆规则，不只是一次性安装说明。如果这个平台支持 skill 或自定义指令，请添加 Memcore Cloud Zhiyi skill/指令；如果这个平台支持 MCP，请注册名为 yifanchen-zhiyi 的 MCP 工具，地址是 http://127.0.0.1:9851/mcp。
安装完成后，只做一次安全能力检查：
{"query":"capability check","mode":"capability_check"}
先不要召回我的真实记忆。
以后请持续遵守这条规则。凡是我的问题依赖旧上下文，请先调用 zhiyi_recall，默认结合 source_refs 回答；只有我明确需要原文证据时，才请求 raw_excerpt。触发词包括：之前、定论、纠错、边界、忘了、安装/测试/发布状态，以及持续项目里的“下一步/接下来呢/还有吗/然后呢”。默认使用 active 召回：当前窗口/session 优先，然后同项目/同工作区、同工作流/同任务、稳定偏好/工具事实。只有我明确要求更宽视图时，才使用 raw-pool/global。如果显式 window-only 召回提示当前窗口/session 还没绑定，请直接说明这个绑定缺口；不要说没有记忆。如果 skill 已安装但 zhiyi_recall 不可用，请告诉我 MCP/工具连接还没接上，不要凭印象猜。
写代码、安装、同步、发版准备或远端排障前，先做只读的开工前检查：
{"query":"<准备做的事>","mode":"work_preflight"}
用它先判断这件事更像已经做了但忘了、已经做了但接线错了、缺诊断入口，还是真的缺功能。这个结果只是起点；动手前仍然要查仓库、测试、工具和文档。
The installer adds the workflow skill where skills are supported, registers yifanchen-zhiyi MCP where the platform supports MCP, and keeps backup/receipt records for local config writes.
macOS / Linux:
curl -fL -o memcore-cloud-install.sh https://github.com/strmforge/memcore-cloud/releases/download/v2026.6.16/install.sh
bash memcore-cloud-install.shWindows PowerShell:
iwr https://github.com/strmforge/memcore-cloud/releases/download/v2026.6.16/install.ps1 -OutFile .\install.ps1
.\install.ps1If you downloaded the release zip, Windows can also use the double-click
Memcore Cloud Installer.cmd; it opens a folder picker and then runs the same
installer with the selected path. On macOS, double-click
Memcore Cloud Installer.command from the extracted release folder.
Windows installs default to %LOCALAPPDATA%\memcore-cloud. To choose a path
before the install:
$env:MEMCORE_INSTALL_DIR = "D:\Apps\memcore-cloud"
iwr https://github.com/strmforge/memcore-cloud/releases/download/v2026.6.16/install.ps1 -OutFile .\install.ps1
.\install.ps1If you already downloaded the repo, you can also run:
.\install.ps1 -Dir "D:\Apps\memcore-cloud"WSL is only for development or advanced testing. Normal Windows installs should use the Windows PowerShell command above.
On Windows, use the Memcore Cloud tray icon after install. On macOS, use the Memcore Cloud menu bar icon. Both can open the local console, show health, and catch up missed records.
You can also open the local console directly:
http://127.0.0.1:9850
For install checks, do not use /zhiyi first. It may run real recall. Ask the client to call zhiyi_recall with:
{"query":"capability check","mode":"capability_check"}A good first result should include:
read_only: true
recall_performed: false
raw_excerpt_returned: false
mcp_tools: ["zhiyi_recall"]
Only run real recall after you explicitly choose to test memory retrieval.
To check whether records are guarded before testing recall, run:
python3 tools/record_doctor.pyIt prints a short read-only report for source records, raw mirrors, the canonical index, and memory/experience links. It does not run recall, backfill, model calls, or platform writes.
Open http://127.0.0.1:9850 to see:
- which AI tools are present on this machine;
- which ones can run a safe capability check;
- which ones are already connected or ready for local AI tool integration;
- whether source records, raw mirrors, the canonical index, and memory/experience links are guarded;
- whether a tool looks recently used or has been quiet for a while;
- where new raw records are being stored.
On Windows and macOS, the tray/menu bar icon gives you the same entry point without remembering the port. The local watcher keeps running and can backfill missed records after restart or repair.
Supported local AI tool entries can be connected automatically. Conversation import uses verified local formats, and capability check remains no-recall until an agent calls real recall.
- Source-backed memory: recall can carry
source_refs, raw excerpts, library ids, and rank reasons. - Zhiyi and Xingce: Zhiyi keeps preference and intent experience; Xingce keeps work experience and validation paths. Experience is not a skill library.
- Read-only pre-work checks: agents can check existing context before they edit, so a finished feature does not get rebuilt just because the next window forgot it.
- Traceable experience evolution: candidates, review queues, validation receipts, apply gates, adoption receipts, and rollback overlays keep useful work paths improving while preserving raw records and receipts.
- Record doctor: a one-click self-check shows whether source records, raw mirrors, the canonical index, and memory/experience links are guarded.
- A timeline you can trace back: different tools leave different clues, but Memcore Cloud keeps them in one source-backed timeline. Raw records stay first; useful experience can settle into Zhiyi, Xingce, toolbook, or errata with source refs, collection ids, lifecycle state, and receipts.
- Organized local records: new records are grouped by computer first, then by the AI tool that produced them, so a multi-device setup can stay understandable.
- Claude is handled carefully: Claude Desktop and Claude Code CLI can both connect, but they remain separate surfaces. Official, relay, and CLI-related records keep attribution boundaries.
- Hermes can inspect sources itself: Memcore Cloud can provide raw/source-ref pointers and observe native feedback, while Hermes-owned skill changes remain Hermes-owned.
2026.6.16 is the current stable release. It focuses on local AI tool connection, pre-work context checks, Record Doctor, source-backed recall, and the candidate, validation, adoption, and rollback path for reusable experience.
See RELEASE_NOTES_2026.6.16.md for this release, UPDATE_HISTORY.md for older highlights, and CHANGELOG.md for lower-level changes.
- Claude Desktop: can use Memcore Cloud through local MCP / Desktop Extensions; source records use verified local format collectors.
- Claude Code CLI: can use MCP while staying separate from Claude Desktop.
- Codex: can use the shared skill and MCP entry, and local sessions can become source-backed records.
- OpenClaw: can receive memory support through its normal local entry points.
- Hermes: can consume raw/source-ref pointers and produce native feedback without Memcore Cloud writing Hermes skills.
- Other local AI tools: can be recognized from local settings, app folders, package managers, and workspace markers; supported local entries can be connected automatically, and tools are promoted to memory sources once their local formats are verified.
macOS / Linux:
~/.memcore-cloud/uninstall.shWindows:
.\uninstall.ps1Uninstalling removes the app files only. Local data such as memory/, raw/, zhiyi/, and config/ is kept.