All Frameworks
CLI Quickstart
As simple as git. Install the agmry CLI and give your agent persistent memory in under 2 minutes.
1Install
$ npm install -g agmry
2Authenticate
Save your API key. Choose project-local (.agmry/) or global (~/.agmry/).
# Option A: interactive login
$ agmry login --local
API key (amk_...): amk_your_key_here
✓ Key saved to .agmry/config.json
# Option B: environment variable
$ export AGMRY_API_KEY="amk_your_key_here"
# Option C: sign up directly (agent-first onboarding)
$ agmry signup my-project --email you@company.com --local
3Use It
Your agent can now store and recall memory from the terminal.
# Store messages
$ agmry store SPACE_ID "What is the capital of France?"
✓ Stored user message
$ agmry store SPACE_ID "The capital of France is Paris." -r assistant
✓ Stored assistant message
# Recall the last 5 messages
$ agmry recall SPACE_ID -n 5
1 user: What is the capital of France?
2 assistant: The capital of France is Paris.
# Search by meaning
$ agmry search SPACE_ID "capital"
[0.92] The capital of France is Paris.
# Set context (key-value)
$ agmry ctx SPACE_ID set preferred_language "English"
✓ Set context "preferred_language"
# Load everything at session start
$ agmry boot SPACE_ID --json
# Full context payload: messages, entries, context, scratchpad
# JSON output for piping into scripts
$ agmry recall SPACE_ID --json | jq '.[].content'
4Add to Agent Instructions
Drop this into your CLAUDE.md, .cursorrules, .clinerules, or .windsurfrules:
## Agentic Memory
Use the agmry CLI for persistent memory across sessions.
Key is in .agmry/config.json (auto-loaded).
agmry store SPACE "what happened" # remember something
agmry recall SPACE # what was said recently
agmry search SPACE "the deadline" # find past context
agmry ctx SPACE set key "value" # store structured data
agmry boot SPACE --json # load everything at session start
5All Commands
$ agmry --help
Run agmry --help for the full reference. Every command supports --json for machine-readable output.
Starter ($1 one-time): 1 space, 1K messages/month. Upgrade to Pro ($24.99/mo) for semantic search + auto-summaries.