memnest
Local memory service that lets coding AIs save and look up past decisions from original text.
Tested on LinuxKnown vulnerabilities
Use it
Before you install: known vulnerabilities
pickfit found 2 known high or critical vulnerabilities in the packages a fresh install brings (OSV.dev, 3 Nov 2026). Weigh this before you install it.
crates:quinn-proto 0.11.15Peer minimum ACK delay can panic quinn-proto when ACK_FREQUENCY transmission is enabled (RUSTSEC-2026-0340, high, fixed in 0.11.18)
See 1 more
crates:quinn-proto 0.11.15— RUSTSEC-2026-0341 (high), fixed in 0.11.17
Linux x86_64
You need first
- Runtimes:
Python 3 is needed for setup - At least 1.9 GB of RAM
download memnest v0.4.3 release archive
curl -fsSL https://github.com/Blue-B/memnest/releases/download/v0.4.3/memnest-v0.4.3-x86_64-unknown-linux-gnu.tar.gz -o /tmp/memnest.tar.gzextract archive and install memnest binary to ~/.local/bin
mkdir -p /tmp/mn_extract && tar -xzf /tmp/memnest.tar.gz -C /tmp/mn_extract && mkdir -p ~/.local/bin && cp /tmp/mn_extract/memnest ~/.local/bin/memnest && chmod +x ~/.local/bin/memnest
Where the README and reality differ
- README: memory_search(query="...")
What worked: requires project or cwd; use project="all" for cross-project search e.g. {"query":"test","project":"all"}
Add it to your agent
claude mcp add memnest -- memnest --mcpJSON for Claude Desktop, Cursor and other clients
{
"mcpServers": {
"memnest": {
"command": "memnest",
"args": [
"--mcp"
]
}
}
}Add memnest as an MCP server in your coding AI's settings, then use its memory tools such as memory_search and memory_remember from your chat.
What it does
These come from the README; pickfit didn’t check them one by one.
- Look up past decisions and the reasons behind themNot tested
- Save important decisions on purposeNot tested
- Share one memory across your coding AIsNot tested
How far pickfit got
Linux x86_64
- Installs
- Connects
- A tool call works
- Works offline
macOS: the project says it runs there; pickfit hasn’t tried it.
Good for
- Find the discussion behind switching task managers
- Look up last settled deployment approach and why
- Look up what was tried last time an error happened
How pickfit tested it
Each run starts from a clean machine and follows the steps above. How pickfit verifies
Linux x86_64
An MCP client connects and lists its tools
- Replayed from scratch
- Works offline
- Checked on 31 Oct 2026
- Version 0.4.3
An MCP client handshake over stdio listed 5 tools.
Differences the test machine hit
- README: curl -fsSL https://raw.githubusercontent.com/Blue-B/memnest/main/core/scripts/install.sh | bash installs and starts service
In the test: install.sh requires systemd-analyze and systemd unit paths; in sandbox it fails with 'systemd-analyze: command not found'. Manually download tarball and copy memnest binary to ~/.local/bin works; run MCP via memnest --mcp - README: first write or search downloads 1.1GB embedding model
In the test: first memory_search took ~18s to download model; subsequent offline calls succeed from cache