agent-memory-systems

Name: agent-memory-systems
Author: sickn33

44kstars6.5kforks

How to install agent-memory-systems?

sickn33/agentic-awesome-skills/agent-memory-systems

$npx -y skills add sickn33/agentic-awesome-skills --skill agent-memory-systems

Installs into the current project.

Summary

Memory is the cornerstone of intelligent agents. Without it, every

Files · 1

View on GitHub

SKILL.md

1	# Agent Memory Systems
2
3	Memory is the cornerstone of intelligent agents. Without it, every interaction
4	starts from zero. This skill covers the architecture of agent memory: short-term
5	(context window), long-term (vector stores), and the cognitive architectures
6	that organize them.
7
8	Key insight: Memory isn't just storage - it's retrieval. A million stored facts
9	mean nothing if you can't find the right one. Chunking, embedding, and retrieval
10	strategies determine whether your agent remembers or forgets.
11
12	The field is fragmented with inconsistent terminology. We use the CoALA cognitive
13	architecture framework: semantic memory (facts), episodic memory (experiences),
14	and procedural memory (how-to knowledge).
15
16	## Principles
17
18	- Memory quality = retrieval quality, not storage quantity
19	- Chunk for retrieval, not for storage
20	- Context isolation is the enemy of memory
21	- Right memory type for right information
22	- Decay old memories - not everything should be forever
23	- Test retrieval accuracy before production
24	- Background memory formation beats real-time
25
26	## Capabilities
27
28	- agent-memory
29	- long-term-memory
30	- short-term-memory
31	- working-memory
32	- episodic-memory
33	- semantic-memory
34	- procedural-memory
35	- memory-retrieval
36	- memory-formation
37	- memory-decay
38
39	## Scope
40
41	- vector-database-operations → data-engineer
42	- rag-pipeline-architecture → llm-architect
43	- embedding-model-selection → ml-engineer
44	- knowledge-graph-design → knowledge-engineer
45
46	## Tooling
47
48	### Memory_frameworks
49
50	- LangMem (LangChain) - When: LangGraph agents with persistent memory Note: Semantic, episodic, procedural memory types
51	- MemGPT / Letta - When: Virtual context management, OS-style memory Note: Hierarchical memory tiers, automatic paging
52	- Mem0 - When: User memory layer for personalization Note: Designed for user preferences and history
53
54	### Vector_stores
55
56	- Pinecone - When: Managed, enterprise-scale (billions of vectors) Note: Best query performance, highest cost
57	- Qdrant - When: Complex metadata filtering, open-source Note: Rust-based, excellent filtering
58	- Weaviate - When: Hybrid search, knowledge graph features Note: GraphQL interface, good for relationships
59	- ChromaDB - When: Prototyping, small/medium apps Note: Developer-friendly, ~20ms p50 at 100K vectors
60	- pgvector - When: Already using PostgreSQL, simpler setup Note: Good for <1M vectors, familiar tooling
61
62	### Embedding_models
63
64	- OpenAI text-embedding-3-large - When: Best quality, 3072 dimensions Note: $0.13/1M tokens
65	- OpenAI text-embedding-3-small - When: Good balance, 1536 dimensions Note: $0.02/1M tokens, 5x cheaper
66	- nomic-embed-text-v1.5 - When: Open-source, local deployment Note: 768 dimensions, good quality
67	- all-MiniLM-L6-v2 - When: Lightweight, fast local embedding Note: 384 dimensions, lowest latency
68
69	## Patterns
70
71	### Memory Type Architecture
72
73	Choosing the right memory type for different information
74
75	When to use: Designing agent memory system
76
77	# MEMORY TYPE ARCHITECTURE (CoALA Framework):
78
79	"""
80	Three memory types for different purposes:
81
82	1. Semantic Memory: Facts and knowledge
83	- What you know about the world
84	- User preferences, domain knowledge
85	- Stored in profiles (structured) or collections (unstructured)
86
87	2. Episodic Memory: Experiences and events
88	- What happened (timestamped events)
89	- Past conversations, task outcomes
90	- Used for learning from experience
91
92	3. Procedural Memory: How to do things
93	- Rules, skills, workflows
94	- Often implemented as few-shot examples
95	- "How did I solve this before?"
96	"""
97
98	## LangMem Implementation
99	"""
100	from langmem import MemoryStore
101	from langgraph.graph import StateGraph
102
103	# Initialize memory store
104	memory = MemoryStore(
105	connection_string=os.environ["POSTGRES_URL"]
106	)
107
108	# Semantic memory: user profile
109	await memory.semantic.upsert(
110	namespace="user_profile",
111	key=user_id,
112	content={
113	"name": "Alice",
114	"preferences": ["dark mode", "concise responses"],
115	"expertise_level": "developer",
116	}
117	)
118
119	# Episodic memory: past interaction
120	await memory.episodic.add(
121	namespace="conversations",
122	content={
123	"timestamp": datetime.now(),
124	"summary": "Helped debug authentication issue",
125	"outcome": "resolved",
126	"key_insights": ["Token expiry was root cause"],
127	},
128	metadata={"user_id": user_id, "topic": "debugging"}
129	)
130
131	# Procedural memory: learned pattern
132	await memory.procedural.add(
133	namespace="skills",
134	content={
135	"task_type": "debug_auth",
136	"steps": ["Check token expiry", "Verify refresh flow"],
137	"example_interaction": few_shot_example,
138	}
139	)
140	"""
141
142	## Memory Retrie

People who liked this also viewed

6 picks

Type

paperclip-create-agentCreate new agents in Paperclip with governance-aware hiring. Use when you need to inspect adapter configuration options, compare existing agent configs, draft a new agent prompt/config, and submit a hSkillsJul 2026157k5
subagent-driven-developmentUse when executing implementation plans with independent tasks in the current sessionSkillsJul 2026154k259k
dispatching-parallel-agentsUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependenciesSkillsJul 2026143k259k
browser-act-skill-forgeForges reusable Skill packages (SKILL.md + scripts) from website exploration via browser-act — no re-exploration later. Use when: user wants a reusable Skill for any website, needs to understand a sitSkillsJul 202685k4.6k
wizardGenerate an interactive bash wizard that walks a human through a manual procedure — third-party setup, a one-off migration, an A→B state transition — opening URLs, capturing values, confirming each stSkillsJul 202675k182k
wonda-cliUsing the Wonda CLI to generate images, videos, music, and audio from the terminal — plus LinkedIn, Reddit, and X/Twitter research and automationSkillsJul 202666k137

agent-memory-systems

44kstars6.5kforks

View on GitHub

How to install agent-memory-systems?

sickn33/agentic-awesome-skills/agent-memory-systems

$npx -y skills add sickn33/agentic-awesome-skills --skill agent-memory-systems

Installs into the current project.

Summary

Memory is the cornerstone of intelligent agents. Without it, every

Files · 1

View on GitHub

SKILL.md

1	# Agent Memory Systems
2
3	Memory is the cornerstone of intelligent agents. Without it, every interaction
4	starts from zero. This skill covers the architecture of agent memory: short-term
5	(context window), long-term (vector stores), and the cognitive architectures
6	that organize them.
7
8	Key insight: Memory isn't just storage - it's retrieval. A million stored facts
9	mean nothing if you can't find the right one. Chunking, embedding, and retrieval
10	strategies determine whether your agent remembers or forgets.
11
12	The field is fragmented with inconsistent terminology. We use the CoALA cognitive
13	architecture framework: semantic memory (facts), episodic memory (experiences),
14	and procedural memory (how-to knowledge).
15
16	## Principles
17
18	- Memory quality = retrieval quality, not storage quantity
19	- Chunk for retrieval, not for storage
20	- Context isolation is the enemy of memory
21	- Right memory type for right information
22	- Decay old memories - not everything should be forever
23	- Test retrieval accuracy before production
24	- Background memory formation beats real-time
25
26	## Capabilities
27
28	- agent-memory
29	- long-term-memory
30	- short-term-memory
31	- working-memory
32	- episodic-memory
33	- semantic-memory
34	- procedural-memory
35	- memory-retrieval
36	- memory-formation
37	- memory-decay
38
39	## Scope
40
41	- vector-database-operations → data-engineer
42	- rag-pipeline-architecture → llm-architect
43	- embedding-model-selection → ml-engineer
44	- knowledge-graph-design → knowledge-engineer
45
46	## Tooling
47
48	### Memory_frameworks
49
50	- LangMem (LangChain) - When: LangGraph agents with persistent memory Note: Semantic, episodic, procedural memory types
51	- MemGPT / Letta - When: Virtual context management, OS-style memory Note: Hierarchical memory tiers, automatic paging
52	- Mem0 - When: User memory layer for personalization Note: Designed for user preferences and history
53
54	### Vector_stores
55
56	- Pinecone - When: Managed, enterprise-scale (billions of vectors) Note: Best query performance, highest cost
57	- Qdrant - When: Complex metadata filtering, open-source Note: Rust-based, excellent filtering
58	- Weaviate - When: Hybrid search, knowledge graph features Note: GraphQL interface, good for relationships
59	- ChromaDB - When: Prototyping, small/medium apps Note: Developer-friendly, ~20ms p50 at 100K vectors
60	- pgvector - When: Already using PostgreSQL, simpler setup Note: Good for <1M vectors, familiar tooling
61
62	### Embedding_models
63
64	- OpenAI text-embedding-3-large - When: Best quality, 3072 dimensions Note: $0.13/1M tokens
65	- OpenAI text-embedding-3-small - When: Good balance, 1536 dimensions Note: $0.02/1M tokens, 5x cheaper
66	- nomic-embed-text-v1.5 - When: Open-source, local deployment Note: 768 dimensions, good quality
67	- all-MiniLM-L6-v2 - When: Lightweight, fast local embedding Note: 384 dimensions, lowest latency
68
69	## Patterns
70
71	### Memory Type Architecture
72
73	Choosing the right memory type for different information
74
75	When to use: Designing agent memory system
76
77	# MEMORY TYPE ARCHITECTURE (CoALA Framework):
78
79	"""
80	Three memory types for different purposes:
81
82	1. Semantic Memory: Facts and knowledge
83	- What you know about the world
84	- User preferences, domain knowledge
85	- Stored in profiles (structured) or collections (unstructured)
86
87	2. Episodic Memory: Experiences and events
88	- What happened (timestamped events)
89	- Past conversations, task outcomes
90	- Used for learning from experience
91
92	3. Procedural Memory: How to do things
93	- Rules, skills, workflows
94	- Often implemented as few-shot examples
95	- "How did I solve this before?"
96	"""
97
98	## LangMem Implementation
99	"""
100	from langmem import MemoryStore
101	from langgraph.graph import StateGraph
102
103	# Initialize memory store
104	memory = MemoryStore(
105	connection_string=os.environ["POSTGRES_URL"]
106	)
107
108	# Semantic memory: user profile
109	await memory.semantic.upsert(
110	namespace="user_profile",
111	key=user_id,
112	content={
113	"name": "Alice",
114	"preferences": ["dark mode", "concise responses"],
115	"expertise_level": "developer",
116	}
117	)
118
119	# Episodic memory: past interaction
120	await memory.episodic.add(
121	namespace="conversations",
122	content={
123	"timestamp": datetime.now(),
124	"summary": "Helped debug authentication issue",
125	"outcome": "resolved",
126	"key_insights": ["Token expiry was root cause"],
127	},
128	metadata={"user_id": user_id, "topic": "debugging"}
129	)
130
131	# Procedural memory: learned pattern
132	await memory.procedural.add(
133	namespace="skills",
134	content={
135	"task_type": "debug_auth",
136	"steps": ["Check token expiry", "Verify refresh flow"],
137	"example_interaction": few_shot_example,
138	}
139	)
140	"""
141
142	## Memory Retrie

People who liked this also viewed

6 picks

Type