Skip to content

Commit aaddaeb

Browse files
committed
feat(17-ai-agentic): add agent-memory, memory-compaction, sub-agent-isolation
1 parent 1bd45ff commit aaddaeb

9 files changed

Lines changed: 3665 additions & 12 deletions

File tree

‎.ref-cache.json‎

Lines changed: 19 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -49,6 +49,7 @@
4949
"https://arxiv.org/abs/2309.15217": 200,
5050
"https://arxiv.org/abs/2310.03714": 200,
5151
"https://arxiv.org/abs/2310.04406": 200,
52+
"https://arxiv.org/abs/2310.08560": 200,
5253
"https://arxiv.org/abs/2310.11511": 200,
5354
"https://arxiv.org/abs/2312.04511": 200,
5455
"https://arxiv.org/abs/2312.10997": 200,
@@ -75,9 +76,12 @@
7576
"https://camel.apache.org/manual/faq/what-is-camel.html": 200,
7677
"https://cassandra.apache.org/doc/4.1/cassandra/cql/mvs.html": 200,
7778
"https://clang.llvm.org/docs/RAVFrontendAction.html": 200,
79+
"https://code.claude.com/docs/en/context-window": 200,
80+
"https://code.claude.com/docs/en/costs": 200,
7881
"https://code.claude.com/docs/en/sub-agents": 200,
7982
"https://code.visualstudio.com/api/extension-guides/command": 200,
8083
"https://code.visualstudio.com/docs/copilot/chat/mcp-servers": 200,
84+
"https://cognition.com/blog/dont-build-multi-agents": 200,
8185
"https://condor.depaul.edu/dmumaugh/OOT/Design-Principles/acv.pdf": 200,
8286
"https://cqrs.files.wordpress.com/2010/11/cqrs_documents.pdf": 200,
8387
"https://cqrs.wordpress.com/wp-content/uploads/2010/11/cqrs_documents.pdf": 200,
@@ -220,8 +224,13 @@
220224
"https://docs.langchain.com/oss/python/integrations/splitters/recursive_text_splitter": 200,
221225
"https://docs.langchain.com/oss/python/langchain/rag": 200,
222226
"https://docs.langchain.com/oss/python/langchain/structured-output": 200,
227+
"https://docs.langchain.com/oss/python/langgraph/add-memory": 200,
223228
"https://docs.langchain.com/oss/python/langgraph/graph-api": 200,
229+
"https://docs.langchain.com/oss/python/langgraph/memory": 200,
224230
"https://docs.langchain.com/oss/python/langgraph/workflows-agents": 200,
231+
"https://docs.letta.com/guides/agents/memory": 200,
232+
"https://docs.letta.com/memory": 200,
233+
"https://docs.letta.com/v1-sdk/messages/compaction": 200,
225234
"https://docs.nginx.com/nginx/admin-guide/security-controls/terminating-ssl-http/": 200,
226235
"https://docs.opensearch.org/latest/query-dsl/compound/hybrid/": 200,
227236
"https://docs.oracle.com/en/java/javase/17/docs/api/java.base/java/lang/Runnable.html": 200,
@@ -312,6 +321,7 @@
312321
"https://download.eclipse.org/microprofile/microprofile-lra-2.0/microprofile-lra-spec-2.0.html": 200,
313322
"https://dspy.ai/api/modules/MultiChainComparison/": 200,
314323
"https://dspy.ai/learn/programming/7-assertions/": 200,
324+
"https://en.wikipedia.org/wiki/Actor_model": 200,
315325
"https://en.wikipedia.org/wiki/Adapter_pattern": 200,
316326
"https://en.wikipedia.org/wiki/Akamai_Technologies": 200,
317327
"https://en.wikipedia.org/wiki/Bastion_host": 200,
@@ -560,6 +570,7 @@
560570
"https://martinfowler.com/eaaDev/EventAggregator.html": 200,
561571
"https://martinfowler.com/eaaDev/EventSourcing.html": 200,
562572
"https://materialize.com/docs/sql/create-materialized-view/": 200,
573+
"https://mem0.ai": 200,
563574
"https://microservices.io/patterns/apigateway.html": 200,
564575
"https://microservices.io/patterns/data/api-composition.html": 200,
565576
"https://microservices.io/patterns/data/cqrs.html": 200,
@@ -585,9 +596,13 @@
585596
"https://nvlpubs.nist.gov/nistpubs/SpecialPublications/NIST.SP.800-88r1.pdf": 200,
586597
"https://open-policy-agent.github.io/gatekeeper/website/docs/": 200,
587598
"https://openai.com/index/introducing-gpt-5/": 200,
599+
"https://openai.github.io/openai-agents-python/handoffs/": 200,
588600
"https://openai.github.io/openai-agents-python/mcp/": 200,
601+
"https://openai.github.io/openai-agents-python/multi_agent/": 200,
602+
"https://openai.github.io/openai-agents-python/tools/": 200,
589603
"https://openrouter.ai/openrouter/auto": 200,
590604
"https://opentelemetry.io/docs/concepts/signals/traces/": 200,
605+
"https://owasp.org/www-project-top-10-for-large-language-model-applications/": 200,
591606
"https://peps.python.org/pep-0008/": 200,
592607
"https://peps.python.org/pep-0008/#comments": 200,
593608
"https://peps.python.org/pep-0318/": 200,
@@ -601,6 +616,7 @@
601616
"https://pkg.go.dev/sort": 200,
602617
"https://pkg.go.dev/sync": 200,
603618
"https://pkg.go.dev/sync#Map.Range": 200,
619+
"https://platform.claude.com/docs/en/agents-and-tools/tool-use/memory-tool": 200,
604620
"https://platform.claude.com/docs/en/agents-and-tools/tool-use/overview": 200,
605621
"https://platform.claude.com/docs/en/agents-and-tools/tool-use/strict-tool-use": 200,
606622
"https://platform.claude.com/docs/en/build-with-claude/extended-thinking": 200,
@@ -674,6 +690,7 @@
674690
"https://wikitech.wikimedia.org/wiki/MediaWiki_Engineering/Guides/Backend_performance_practices": 200,
675691
"https://www.abebooks.com/9781621140368/Cloud-Design-Patterns-Prescriptive-Architecture-1621140369/plp": 200,
676692
"https://www.anthropic.com/engineering/building-effective-agents": 200,
693+
"https://www.anthropic.com/engineering/effective-context-engineering-for-ai-agents": 200,
677694
"https://www.anthropic.com/engineering/multi-agent-research-system": 200,
678695
"https://www.anthropic.com/news/contextual-retrieval": 200,
679696
"https://www.anthropic.com/news/model-context-protocol": 200,
@@ -728,6 +745,7 @@
728745
"https://www.kernel.org/doc/html/latest/process/coding-style.html": 200,
729746
"https://www.langchain.com/blog/agentic-rag-with-langgraph": 200,
730747
"https://www.langchain.com/blog/planning-agents": 200,
748+
"https://www.letta.com/blog/memgpt-and-letta": 200,
731749
"https://www.microsoft.com/en-us/research/articles/magentic-one-a-generalist-multi-agent-system-for-solving-complex-tasks/": 200,
732750
"https://www.mongodb.com/docs/manual/core/sharding-shard-key/": 200,
733751
"https://www.mongodb.com/docs/manual/indexes/": 200,
@@ -759,6 +777,7 @@
759777
"https://www.semanticscholar.org/paper/Reflexion:-an-autonomous-agent-with-dynamic-memory-Shinn-Labash/46299fee72ca833337b3882ae1d8316f44b32b3c": 202,
760778
"https://www.slf4j.org/manual.html": 200,
761779
"https://www.thoughtworks.com/en-us/insights/articles/embracing-strangler-fig-pattern-legacy-modernization-part-one": 200,
780+
"https://www.trychroma.com/research/context-rot": 200,
762781
"https://www.uml-diagrams.org/state-machine-diagrams.html": 200,
763782
"https://www.usenix.org/conference/hotcloud16/workshop-program/presentation/burns": 200,
764783
"https://www.usenix.org/conference/nsdi13/technical-sessions/presentation/nishtala": 200,

‎README.md‎

Lines changed: 2 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -5,7 +5,7 @@ primary sources, carries eighteen mandatory dimensions, and cites every claim.
55

66
![License](https://img.shields.io/badge/license-CC%20BY%204.0-blue)
77
![Families](https://img.shields.io/badge/families-29-informational)
8-
![Entries](https://img.shields.io/badge/entries-95%20published%20%2F%20887%20planned-yellow)
8+
![Entries](https://img.shields.io/badge/entries-98%20published%20%2F%20887%20planned-yellow)
99
![Dimensions per entry](https://img.shields.io/badge/dimensions%20per%20entry-18-green)
1010
![Citations](https://img.shields.io/badge/citations-verified%20in%20CI-brightgreen)
1111
![Original prose](https://img.shields.io/badge/prose-100%25%20original-brightgreen)
@@ -90,7 +90,7 @@ README does not claim they are implemented.
9090
| 14 | [Testing](patterns/14-testing/) | Meszaros, xUnit Test Patterns | 0 | 30 | 30 |
9191
| 15 | [Security](patterns/15-security/) | OWASP ASVS | 0 | 38 | 38 |
9292
| 16 | [Functional Programming](patterns/16-functional/) | Category theory in practice | 0 | 40 | 40 |
93-
| 17 | [AI and Agentic](patterns/17-ai-agentic/) | Papers and vendor engineering, 2023 to 2026 | 27 | 28 | 55 |
93+
| 17 | [AI and Agentic](patterns/17-ai-agentic/) | Papers and vendor engineering, 2023 to 2026 | 30 | 25 | 55 |
9494
| 18 | [Anti-Patterns](patterns/18-anti-patterns/) | Brown et al, AntiPatterns | 0 | 53 | 53 |
9595
| 19 | [API and Interface Design](patterns/19-api-design/) | REST, GraphQL, gRPC specifications | 0 | 10 | 10 |
9696
| 20 | [Release and Deployment](patterns/20-release-deployment/) | Humble and Farley | 0 | 10 | 10 |

‎dist/catalogue-status.csv‎

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -15,7 +15,7 @@ family,name,published,planned,target,percent
1515
14-testing,Testing,0,30,30,0.0
1616
15-security,Security,0,38,38,0.0
1717
16-functional,Functional Programming,0,40,40,0.0
18-
17-ai-agentic,AI and Agentic,27,28,55,49.1
18+
17-ai-agentic,AI and Agentic,30,25,55,54.5
1919
18-anti-patterns,Anti-Patterns,0,53,53,0.0
2020
19-api-design,API and Interface Design,0,10,10,0.0
2121
20-release-deployment,Release and Deployment,0,10,10,0.0

‎dist/catalogue-status.json‎

Lines changed: 5 additions & 5 deletions
Original file line numberDiff line numberDiff line change
@@ -1,6 +1,6 @@
11
{
22
"generated_by": "tools/gen-catalogue-status.py",
3-
"published_total": 95,
3+
"published_total": 98,
44
"target_total": 887,
55
"families_total": 29,
66
"families_complete": 2,
@@ -177,12 +177,12 @@
177177
"family": "17-ai-agentic",
178178
"name": "AI and Agentic",
179179
"origin": "Papers and vendor engineering, 2023 to 2026",
180-
"published": 27,
181-
"planned": 28,
180+
"published": 30,
181+
"planned": 25,
182182
"target": 55,
183-
"percent": 49.1,
183+
"percent": 54.5,
184184
"maturity": {
185-
"established": 13,
185+
"established": 16,
186186
"emerging": 5,
187187
"canonical": 9
188188
}

‎docs/PROGRESS.md‎

Lines changed: 3 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -3,9 +3,9 @@
33
Generated by `tools/gen-catalogue-status.py`. Never hand-edit; CI
44
checks this file stays current with real repository state.
55

6-
Published: 95
6+
Published: 98
77
Target (published plus planned): 887
8-
Completion: 10.7%
8+
Completion: 11.0%
99
Families: 29
1010
Families complete: 2
1111

@@ -27,7 +27,7 @@ Families complete: 2
2727
| 14 | Testing | 0 | 30 | 30 | 0.0% |
2828
| 15 | Security | 0 | 38 | 38 | 0.0% |
2929
| 16 | Functional Programming | 0 | 40 | 40 | 0.0% |
30-
| 17 | AI and Agentic | 27 | 28 | 55 | 49.1% |
30+
| 17 | AI and Agentic | 30 | 25 | 55 | 54.5% |
3131
| 18 | Anti-Patterns | 0 | 53 | 53 | 0.0% |
3232
| 19 | API and Interface Design | 0 | 10 | 10 | 0.0% |
3333
| 20 | Release and Deployment | 0 | 10 | 10 | 0.0% |

‎patterns/17-ai-agentic/README.md‎

Lines changed: 9 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -2,7 +2,7 @@
22

33
Origin. Papers and vendor engineering, 2023 to 2026
44

5-
27 entries, 233,147 words. Every entry carries all 18
5+
30 entries, 260,192 words. Every entry carries all 18
66
dimensions from [the entry contract](../../docs/ENTRY-TEMPLATE.md).
77

88
## AI Agentic
@@ -16,6 +16,7 @@ dimensions from [the entry contract](../../docs/ENTRY-TEMPLATE.md).
1616
| [Corrective RAG](corrective-rag.md) | emerging | 8,133 | A support bot answers questions by retrieving from a company's help center index and handing the top few documents to a generator. |
1717
| [GraphRAG](graphrag.md) | emerging | 7,850 | A team has a large private corpus, contracts, incident postmortems, research notes, support transcripts, and wants an LLM to answer questions grounded in that corpus. |
1818
| [Late Chunking](late-chunking.md) | emerging | 8,222 | Picture a retrieval pipeline built over a long, single-narrative source, a Wikipedia article, a signed contract, a meeting transcript, a product manual. |
19+
| [Memory Compaction](memory-compaction.md) | established | 8,946 | An agent that runs for a long time accumulates a conversation. |
1920
| [Parallelization](parallelization.md) | established | 9,204 | An agentic pipeline built as a single sequential chain of LLM calls has one throughput limit, the wall-clock latency of the slowest single call multiplied by the number of calls ... |
2021
| [Reranking](reranking.md) | canonical | 7,225 | A reader who has never heard the word reranking has still hit the problem it solves. |
2122
| [Retrieval Augmented Generation](retrieval-augmented-generation.md) | canonical | 8,731 | A large language model's knowledge is frozen at the moment its training data was collected, and its parameters have a fixed, finite capacity that cannot hold every fact a user ... |
@@ -28,6 +29,12 @@ dimensions from [the entry contract](../../docs/ENTRY-TEMPLATE.md).
2829
|---|---|---|---|
2930
| [Tree of Thoughts](tree-of-thoughts.md) | established | 9,275 | A language model produces its answer as a single left-to-right token stream. |
3031

32+
## AI/Agentic
33+
34+
| Pattern | Maturity | Words | Intent |
35+
|---|---|---|---|
36+
| [Agent Memory](agent-memory.md) | established | 8,301 | An agent that runs a single request and returns an answer does not need this pattern. |
37+
3138
## Agentic
3239

3340
| Pattern | Maturity | Words | Intent |
@@ -51,6 +58,7 @@ dimensions from [the entry contract](../../docs/ENTRY-TEMPLATE.md).
5158
| Pattern | Maturity | Words | Intent |
5259
|---|---|---|---|
5360
| [Orchestrator-Worker](orchestrator-worker.md) | established | 10,095 | A task arrives whose internal shape cannot be known until the model has already looked at it. |
61+
| [Sub-Agent Isolation](sub-agent-isolation.md) | established | 9,798 | An agent delegates a subtask to another agent, and the subtask involves work whose intermediate output the delegating agent will never need again. |
5462

5563
## Reasoning
5664

0 commit comments

Comments
 (0)