A shared memory service for AI agents with two parts: signed geographic measurements and a writable scratchpad for agent notes. Measurements are tied to locations and times, with signatures and labels showing their source or trust level.

Use it to recall, compare, find similar, and manage geographic facts and agent memory across sessions.

It lets agents retrieve shared facts while distinguishing sensor readings, calculated data, model output, and human-curated information. The scratchpad provides a place to store and manage agent-owned memory.

emem — Plugin for Gemini CLI by Vortx-AI

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

Clone the repo

git clone --depth 1 https://github.com/Vortx-AI/emem

Made for: Gemini CLI.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

README.md

[![agentmods](https://agentmods.dev/badge/plugins/vortx-ai/emem/examples.svg)](https://agentmods.dev/plugins/vortx-ai/emem/examples)

Your own site

<a href="https://agentmods.dev/plugins/vortx-ai/emem/examples"><img src="https://agentmods.dev/badge/plugins/vortx-ai/emem/examples.svg" alt="Measured on agentmods" height="20"></a>

Per session not measured What this adds to a session before it is invoked.

When invoked not measured Not applicable: nothing here is loaded into a session.

Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.

Origin original No closer match found in the catalogue.

Security

Grade A, and why

emem scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.

The scan reads gemini-extension.json. This mod also ships 19 executable files (3d-worlds/capture.mjs, 3d-worlds/emem-world.js, 3d-worlds/make_splats.py, …), listed below but not scanned — reading those needs a real analyzer, not pattern matching.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

examples/gemini-extension.json · 68 lines

How it starts

The opening of the file, as written. The whole thing — 68 lines — stays where its author put it; the contents beside it link to each section on GitHub.

{
  "_doc": "Gemini CLI extension manifest. Install with one command: `gemini extensions install https://emem.dev/gemini-extension.json`. After install, every Gemini CLI session can call emem_recall, emem_compare, emem_find_similar, etc. against any place on Earth without an API key. The hosted instance at https://emem.dev/mcp is HTTPS-only Streamable HTTP transport (MCP 2025-03-26+). For self-hosted, change `url` to your own emem responder.",
  "name": "emem",
  "version": "1.1.0",
  "description": "Shared, verifiable memory for AI agents. Two layers, one trust surface. Earth-scale fact substrate: every patch of ground gets a 64-bit address (cell64, about 9.55 m), and every measurement at that cell is a signed fact keyed by (cell, band, tslot), ed25519-signed by the responder over the blake3 of its canonical CBOR, with a per-band tamper-provenance class (direct_sensor, deterministic_index, model_output, human_curated) so an agent can tell a raw sensor reading from a model estimate. Agent memory: above the fact substrate sits the writable scratchpad the agent owns: six Anthropic memory-tool verbs (emem_memory_view/create/str_replace/insert/delete/rename), CoALA-typed files (episodic/semantic/procedural/resource), capability-bound writes under /memories/by_attester/, BGE-768 semantic search over notes, multi-attester contradiction scoring per band kind, signed bundles (emem:bundle:<bundle_cid>), field tokens for areas (emem:raster: is a native-resolution grid over a bbox as one content-addressed signed artifact, emem:cube: is a field over time as a signed manifest over raster slices; enumerate an area's cells with cells_in_bbox), the change-attribution evidence ledger (why a readout moved, per term, split null by design), and bi-temporal recall (as_of_tslot for valid time, as_of_signed_at for transaction time, both threaded through every read). 97 MCP tools (14 core, 83 extended); tools/list at /mcp advertises the core tier and /mcp/full advertises all 97, with every

Read the full file on GitHub · 68 lines

Files

What ships with it

60 files beside gemini-extension.json in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

yesterday First seen · 68 lines scan A 811459d82f09

Subscribe to this mod's changes

emem is a plugin published in the GitHub repository Vortx-AI/emem (56 stars, last pushed yesterday), licensed Apache-2.0. Its token cost is not measured: this kind of file is read by the harness, not the model. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-06.

Memory and context Scientific a2a-protocol agent-interoperability agent-memory long-term-memory mcp shared-memory

memX

memX: local-first semantic memory for coding agents. Native Claude Code lifecycle hooks.

NeoLi00/memX · not measured

pseudolife-memory

Persistent long-term memory for Claude Code — wiring for the Pseudolife-MCP daemon (session briefing + identity hooks + memory commands; MCP transport via the installer's stdio shim or claude mcp add). Requires the daemon stack: https://github.com/Pseudogiant-xr/Pseudolife-MCP#quickstart.

Pseudogiant-xr/Pseudolife-MCP · not measured

memory-layer

Persistent memory with outcome-based learning for AI coding agents. Stores knowledge from coding sessions, learns which memories actually help, and gets smarter over time.

runtimenoteslabs/memory-layer · not measured

deja

deja-vu memory for Gemini CLI: the coding sessions already on your disk — Gemini CLI, Claude Code, Codex, Cursor and eighteen more — searchable and recalled, including work from before it was installed.

vshulcz/deja-vu · not measured

engraphis-memory

Give agents durable, scoped, explainable memory across sessions and repos via the Engraphis MCP tools. Use when you learn something worth keeping, need prior context before acting, or ask why/how a fact changed. Covers remember/recall, why/timeline, forget/pin/correct, sessions, and code search.

Coding-Dev-Tools/engraphis · not measured

causal-memory

Persistent causal memory for Claude Code — decisions, outcomes, and lessons that survive compaction. Same-context counterfactuals (natural experiments), a falsifiable prediction ledger, and an activation skill that teaches when to recall and record.

JingxuanC/causal-memory · not measured