Your First Memory

Add, search, and retrieve memories in 3 minutes — both hosted and open source

This cookbook gets you from zero to working memory in 3 minutes. Choose your version and follow along.

Install

Hosted (lyzr-adk)

pip install lyzr-adk

export LYZR_API_KEY="your-lyzr-api-key"

Open Source (lyzr-cognis)

pip install lyzr-cognis

export GEMINI_API_KEY="your-gemini-key"    # For embeddings
export OPENAI_API_KEY="your-openai-key"    # For fact extraction

Step 1: Initialize and Add Memories

Hosted (lyzr-adk)

from lyzr import Cognis, CognisMessage

cog = Cognis()

# Store a conversation
cog.add(
    messages=[
        CognisMessage(role="user", content="My name is Alice and I love hiking."),
        CognisMessage(role="assistant", content="Nice to meet you, Alice!"),
        CognisMessage(role="user", content="I work at Google as a data scientist."),
        CognisMessage(role="assistant", content="That's a great role!"),
    ],
    owner_id="user_alice",
)

Open Source (lyzr-cognis)

from cognis import Cognis

m = Cognis(owner_id="user_alice")

# Store a conversation
result = m.add([
    {"role": "user", "content": "My name is Alice and I love hiking."},
    {"role": "assistant", "content": "Nice to meet you, Alice!"},
    {"role": "user", "content": "I work at Google as a data scientist."},
    {"role": "assistant", "content": "That's a great role!"},
])
print(result["message"])
# "Extracted 3 memories from 4 messages"

Cognis automatically extracts discrete facts from the conversation and stores them as searchable memory records.

Step 2: Search Memories

Hosted (lyzr-adk)

results = cog.search(query="What does Alice do for work?", owner_id="user_alice", limit=5)
for r in results:
    print(f"  {r.content}  (score: {r.score})")

# Output:
#   Alice works at Google as a data scientist  (score: 0.89)

Open Source (lyzr-cognis)

resp = m.search("What does Alice do for work?", limit=5)
for r in resp["results"]:
    print(f"  {r['content']}  (score: {r['score']})")

# Output:
#   Alice works at Google as a data scientist  (score: 0.8712)

Step 3: Retrieve All Memories

Hosted (lyzr-adk)

memories = cog.get(owner_id="user_alice")
for mem in memories.memories:
    print(f"  [{mem.id[:8]}] {mem.content}")

# Output:
#   [a1b2c3d4] Alice's name is Alice
#   [e5f6g7h8] Alice works at Google as a data scientist
#   [i9j0k1l2] Alice loves hiking

Open Source (lyzr-cognis)

resp = m.get_all()
for mem in resp["memories"]:
    cat = mem["metadata"]["category"]
    print(f"  [{cat}] {mem['content']}")

# Output:
#   [identity] Alice's name is Alice
#   [work_career] Alice works at Google as a data scientist
#   [interests] Alice loves hiking

Step 4: Use Context in Your LLM

Hosted (lyzr-adk)

context = cog.context(
    current_messages=[
        CognisMessage(role="user", content="What should I do this weekend?"),
    ],
    owner_id="user_alice",
)
# Use context in your LLM system prompt
print(context)

Open Source (lyzr-cognis)

ctx = m.get_context(
    messages=[{"role": "user", "content": "What should I do this weekend?"}]
)
# Use context_string in your LLM system prompt
print(ctx["context_string"])
# "Relevant memories:\n- Alice loves hiking\n- ..."

m.close()  # Don't forget to close!

What’s Different Between Hosted and Open Source?

Aspect	Hosted	Open Source
Data storage	Lyzr cloud	Local (~/.cognis/)
Async support	Full (aadd, asearch, etc.)	Sync only
Summaries	Available	Not available
Return types	Objects (.content, .score)	Dicts (["content"], ["score"])
Cleanup	Optional	Required (m.close())

For a full comparison, see Hosted vs Open Source.

Next Steps

CrewAI Integration

Add memory to multi-agent crews

LangGraph Integration

Memory as graph nodes