Skip to content

Agentscreator/Engram

Engram

Shared memory for your team's agents

Persistent memory that survives across sessions and detects when agents contradict each other.

License MCP Python


What It Does

When one agent discovers something important — a hidden side effect, a failed approach, an undocumented constraint — it commits that fact. Every other agent on your team can query it instantly.

When two agents develop incompatible beliefs, Engram detects the contradiction and surfaces it for review.

Your data is private. All data is encrypted, isolated by workspace, and never read, analyzed, or redistributed. We have a deep commitment to privacy.


Quick Start

macOS / Linux:

curl -fsSL https://engram-us.com/install | sh

Windows PowerShell:

irm https://engram-us.com/install.ps1 | iex

Windows CMD:

curl -fsSL https://engram-us.com/install.cmd -o install.cmd && install.cmd && del install.cmd

By default, the installer writes https://mcp.engram.app/mcp into your MCP config. If your environment needs a different endpoint, set ENGRAM_MCP_URL before running the installer.

Restart your editor, then ask your agent:

"Set up Engram for my team"

Your agent handles the rest.


Running Locally

If you want to run Engram from this repository during development:

pip install -e ".[dev]"
python -m engram.cli serve --http

Then open:

http://127.0.0.1:7474/dashboard

If engram is not on your PATH, python -m engram.cli ... works reliably.


Setup Flow

Creating a Workspace

🤖 Agent

Do you have an Invite Key to join an existing workspace, or are you setting up a new one?

👤 You

New

🤖 Agent

✅ Your team workspace is ready.

Share this Invite Key with teammates:

ek_live_abc123...

That's all they need — one key, nothing else to configure.

Should commits show who made them, or stay anonymous?



🎥 Getting Started Video

Watch this quick demo to get started with Engram:

(Video coming soon — placeholder added for future update)

This video walks through installation, setup, and basic usage.


Joining a Workspace

🤖 Agent

Do you have an Invite Key to join an existing workspace, or are you setting up a new one?

👤 You

Join — here's my key: ek_live_abc123...

🤖 Agent

You're in. I'll query team memory before starting work on anything.

That's it. Teammates only need the Invite Key. No database URL, no Team ID, no configuration.


How It Works

┌──────────────────────────────────────────┐
│            MCP Tools                     │
│  engram_commit  — Write a fact           │
│  engram_query   — Read team knowledge    │
│  engram_conflicts — See disagreements    │
│  engram_resolve — Settle conflicts       │
├──────────────────────────────────────────┤
│        Conflict Detection                │
│  Tier 0: Entity exact-match              │
│  Tier 1: NLI cross-encoder (local)       │
│  Tier 2: Numeric/temporal rules          │
│  Tier 3: LLM escalation (rare)           │
├──────────────────────────────────────────┤
│          Hosted Storage                  │
│  Managed Postgres — zero setup           │
│  Isolated per workspace                  │
└──────────────────────────────────────────┘

No database to provision, no servers to run, no ports to open. Install and go.


Privacy & Security

Your memory is yours. This isn't a footnote — it's the foundation Engram is built on.

Encrypted. All data is encrypted in transit (TLS) and at rest. Invite keys use encrypted payloads so teammates never see raw credentials.

Isolated. Every workspace is fully isolated. There is no cross-workspace access, no shared tables, no data leakage between teams.

Never read. We don't read your facts. We don't analyze your memory. We don't train on your data. We don't sell it. We have no analytics pipeline that touches your content. Period.

Never redistributed. Your team's knowledge never leaves your workspace. It is never shared with other users, other teams, or third parties. Not now, not ever.

You control it. Delete your workspace and everything is gone. Anonymous mode strips engineer names from all commits. Anonymous agents randomize agent IDs each session. You decide what's visible and what isn't.


Tools

Tool Purpose
engram_commit Persist a verified discovery
engram_query Pull what your team's agents know
engram_conflicts Surface contradictions
engram_resolve Settle disagreements
engram_promote Graduate ephemeral memory to durable

CLI Commands

engram install              # Auto-detect IDEs and configure MCP
engram serve               # Start MCP server (stdio mode)
engram serve --http        # Start MCP server (HTTP mode)
engram setup              # One-command workspace setup
engram status             # Show workspace status
engram info               # Display detailed workspace info
engram whoami             # Show current user identity
engram search <query>     # Query workspace from terminal
engram stats              # Show workspace statistics
engram config show        # Display configuration
engram config set <key>   # Update configuration
engram tail               # Live stream of workspace commits
engram verify             # Verify installation
engram completion <shell> # Install shell tab completion

Conflict Detection

Runs asynchronously in the background:

Tier Method Catches
0 Entity matching "rate limit is 1000" vs "rate limit is 2000"
1 NLI cross-encoder Semantic contradictions
2 Numeric rules Different values for same entity
3 LLM escalation Ambiguous cases (rare, optional)

Commits return instantly. Detection completes in the background (~2-10s on CPU).


Memory That Forgets on Purpose

Engram doesn't just accumulate — it actively forgets what doesn't earn its place.

  • Ephemeral memory — Scratchpad facts auto-expire in 24h unless queried twice ("proved useful more than once")
  • Importance decay — Unverified inferences expire after 30 days. Unverified observations expire after 90 days.
  • Protected facts — Decisions, verified facts, and corroborated claims are never auto-retired.
  • Steeper recency curve — A 90-day-old fact scores 0.001 in retrieval. Old context stops crowding out what matters now.

Grounded in the FiFA/MaRS research on forgetting-by-design for cognitive agents.


Research Foundation

Engram is grounded in peer-reviewed research on multi-agent memory systems:

Full literature review: docs/LITERATURE.md
Implementation details: docs/IMPLEMENTATION.md


Contributing

PRs welcome. See CONTRIBUTING.md.


License

Apache 2.0


An engram is the physical trace a memory leaves in the brain — the actual unit of stored knowledge.

Releases

No releases published

Packages

 
 
 

Contributors

Languages