Java Client 0.1.0

Changelog

🚀 Features

• 7884b45 upgrading java client to jvm 21 (#149), closes #149
• 5dd30a1 use reusable workflow for Java client release
• 17189a3 jreleaser for the Java Client
• 712b54c js-client: v0.3.1 - add utility methods, fix vulnerabilities, fix body read bug
• d04f8fa config: make extraction debounce TTL configurable (#112), closes #112
• aa3028e llm: use LLMClient for startup model validation
• 0bd0069 llm: add LLMClient abstraction layer with LiteLLM backend
• 5a473a5 initial implementation of the Java SDK Client
• 38f71e9 Add LangChain integration with automatic tool conversion
• 562d897 Add recent_messages_limit parameter and fix critical extraction logic bugs
• 2f4ea93 Add --no-worker CLI option for simplified development
• c8bdeb9 complete client library parameter naming updates
• 9db522b complete vectorstore adapter parameter name updates
• 453c7b5 expand short recency parameter names to descriptive ones
• e447288 redis: AggregateQuery with KNN + APPLY + SORTBY for server_side_recency; formatting fixes
• a6d1961 redis: use RedisVL paging with contextlib.suppress; fix loop var lint; adjust next_offset/total
• 576d6c5 redis: use RangeQuery when distance_threshold is provided; VectorQuery otherwise (RedisVL expects distance_threshold at query level)
• 24d321d redis: DB-level recency ranking via RedisVL VectorQuery and adapter fallbacks; format fixes
• 2d38015 Make OpenAI and Anthropic API base URLs configurable
• a69add0 restructure API URLs for better clarity and consistency
• 5638bff restructure API URLs for better clarity and consistency, closes #17
• 0f59a98 Add comprehensive OAuth2/JWT authentication with extensive test coverage, closes #12

🐛 Fixes

• c52be66 use reusable workflow with git-root-search for Java client release
• cab88b6 add --git-root-search flag for JReleaser (#153), closes #153
• 7ea38a3 add git-root-search for JReleaser in monorepo
• bf9eae7 configure tag prefix for Java client release (#152), closes #152
• 62c2eba configure Axion scmVersion block before version resolution (#151), closes #151
• 0f50bdf correct Axion repository path for Java client release (#150), closes #150
• 4b67664 address PR review comments
• 3aba394 add per-task timeouts for LLM-dependent Docket tasks (#135), closes #135
• 9bcca5f add per-task timeouts for LLM-dependent Docket tasks
• 4b2d704 ci: use NPM_TOKEN for npm publish
• 8e61ca1 update minimum Node.js version to 20 for JS client
• f9e33ec index sessions in sorted set for listing (#119), closes #119
• 59679e0 tests: fix pre-existing test failures
• edcaa57 replace silent fallbacks with explicit ModelValidationError exceptions
• 4f602fe patch correct LLMClient.get_model_config path in test
• d387f24 linting/formatting
• 0ebea67 fix type hints and UTs
• 12accf7 enable auth by default
• 6be3056 AWS credentials requirements
• 5249efd AWS credentials check
• a7793cd bedrock runtime type hint
• 90fd640 detailed error message
• 5da85a2 hardcoded docker image tags
• ffd7b25 some mypy issues
• 612932e unit tests
• 207315a bug
• 5e4e864 formatting
• 43c73e4 app startup
• 745ec70 Accept single memory object in create_long_term_memory tool
• 3c3042a Complete LangChain integration example in README
• 3d17dd9 Address Copilot review comments
• dbbf383 Resolve mypy unused-ignore error for optional langchain import
• 4fc4687 Address CI failures and documentation issues
• 68c5138 Use ThreadPoolExecutor to avoid event loop conflicts in HybridBackgroundTasks
• 325bb71 Correct HybridBackgroundTasks interface and behavior
• be0abca add robust error handling for LLM response parsing
• 9dcc4c1 add robust error handling for LLM response parsing
• c1f0729 update test imports after moving recency functions
• 5f849dc address PR review feedback
• a8fb65c address PR feedback - improve type checking, extract complex logic, update docs
• 4c6b1c1 add _parse_list_field to base adapter; tests now pass including recency and adapter paths
• a633044 redis: coerce list fields in Redis aggregate path; add RecencyAggregationQuery.build_args helper; update tests to use build_args; whitespace fix
• 3df5d15 Use elif instead of if for Anthropic provider check
• 61c7497 Fix threading.Lock isinstance test in test_auth.py
• 1d9544b Address all linting issues in auth.py
• 89235f8 Address test failures and dependency issues
• e5ea85c Address test failures and configuration issues
• 846c2c2 Replace boolean lock with threading.Lock for thread safety

🔄️ Changes

• 968be93 llm: remove TestLLMHandler Protocol and test injection mechanism
• 7e805a1 convert LLM response types from dataclasses to Pydantic BaseModel
• 46d80bc consolidate embeddings creation in LLMClient
• 3a83585 llm: update imports to use llm package
• fc212a0 llm: reorganize llm_client.py into llm/ package
• 883fb2e 🚨 llm: migrate from llms.py to LLMClient - llms.py module removed, use llm_client.py instead
• ed3b250 resolve PR review comments on memory editing functionality
• 58ee06b resolve PR review comments on recency and MCP changes
• a1a5a4d move imports to top of vectorstore_adapter.py module
• 83d5abb improve code quality and remove duplication in vectorstore adapter
• eac6268 redis: integrate RecencyAggregationQuery; fix AggregationQuery usage; clean imports

🧪 Tests

• 37f4dea llm: add integration tests for HuggingFace and Gemini providers
• 5a5014a llm: refactor LLM client tests to use standard mocking patterns
• 8bbbae9 update tests for Pydantic model conversion
• d5081a1 llm: update test imports and patch paths for llm package
• 8e4798a llm: update test imports to use LLMClient
• 71260cc llm: add LLMClient tests and update test fixtures
• 5e1f7c5 redis: add RecencyAggregationQuery and server_side_recency adapter tests; formatting fixes

🧰 Tasks

• 31f8560 rename JS client package to unscoped 'agent-memory-client'
• bc23df0 add dev_docs to .gitignore
• 93254c4 Bump client version to 0.12.5

🛠 Build

• 48fa562 use npm trusted publishing (OIDC) instead of NPM_TOKEN
• bdaa9a9 deps: bump marshmallow from 3.26.1 to 3.26.2 (#128), closes #128
• 6ec7da9 deps: update dependencies for LLMClient

📝 Documentation

• ca10332 update project structure for llm/ package
• 63c3384 update CLAUDE.md with LLMClient context
• 18f1e25 Make all LangChain integration examples runnable
• 9968f7b Improve LangChain integration documentation
• 22c1cde rename Memory Strategies to Memory Extraction Strategies and integrate Memory Editing into Memory Lifecycle
• 26c7932 restructure memory documentation
• 6b8275c fix linting issues in memory-lifecycle.md
• ead73d4 tighten language in memory-lifecycle.md
• aab8d86 rewrite memory-lifecycle.md to reflect actual server-controlled implementation
• 5455792 add descriptive parameter examples

Contributors

We'd like to thank the following people for their contributions:

• Andrew Brookins ()
• Brian Sam-Bodden ()
• Chris Guidry (@chrisguidry)
• Copilot ()
• Epic Flow Worker (@tegument)
• Farouk Boukil ()
• GitHub ()
• Julien Ruaux ()
• Nitin Kanukolanu ()
• Raphael De Lio ()
• Song Luar ()
• abrookins (@abrookins)
• allie.jones ()
• claude[bot] (@claude[bot])
• copilot-swe-agent[bot] ()
• dependabot[bot] (@dependabot[bot])
• epicdigitalmedia (@epicdigitalmedia)
• fcenedes (@fcenedes)
• nitin (@nkanu17)

0.13.2

What's Changed

• Fix: Add per-task timeouts for LLM-dependent Docket tasks by @abrookins in #138
• Fix: Add userId to JS client getWorkingMemory options by @abrookins in #137

Full Changelog: server/v0.13.1...server/v0.13.2

0.13.1

What's Changed

Installation

PyPI:

pip install agent-memory-server==0.13.1

Docker (standard - requires external Redis):

docker pull redislabs/agent-memory-server:0.13.1
# or
docker pull ghcr.io/redis/agent-memory-server:0.13.1

Docker (standalone - all-in-one with embedded Redis):

docker pull redislabs/agent-memory-server:0.13.1-standalone
# or
docker pull ghcr.io/redis/agent-memory-server:0.13.1-standalone

# Run with persistent storage:
docker run -d -v ~/.agent-memory/data:/data -p 8000:8000 \
  -e OPENAI_API_KEY=your-key \
  redislabs/agent-memory-server:0.13.1-standalone

0.13.0

What's Changed

Installation

PyPI:

pip install agent-memory-server==0.13.0

Docker (standard - requires external Redis):

docker pull redislabs/agent-memory-server:0.13.0
# or
docker pull ghcr.io/redis/agent-memory-server:0.13.0

Docker (standalone - all-in-one with embedded Redis):

docker pull redislabs/agent-memory-server:0.13.0-standalone
# or
docker pull ghcr.io/redis/agent-memory-server:0.13.0-standalone

# Run with persistent storage:
docker run -d -v ~/.agent-memory/data:/data -p 8000:8000 \
  -e OPENAI_API_KEY=your-key \
  redislabs/agent-memory-server:0.13.0-standalone

0.12.7

What's Changed

• LiteLLM Embeddings Wrapper and Documentation Updates by @nkanu17 in #108
• Use LLM for topic and entity extraction by default by @abrookins in #111
• • Make extraction debounce TTL configurable by @bsbodden in #112
• Fix deduplication bypass in promote_working_memory_to_long_term by @nkanu17 in #114
• Bug fixes, updating supported model versions, and summary views by @abrookins in #103

New Contributors

• @bsbodden made their first contribution in #112

Full Changelog: server/v0.12.6...server/v0.12.7

0.12.6

What's Changed

• Fix non-Docket background tasks for MCP servers by @abrookins in #101
• Allow customizing memory tool descriptions with ToolSchema wrapper classes by @fcenedes in #100
• Working memory now required to have timestamps by @fcenedes in #102
• Initial implementation of the Java SDK Client by @raphaeldelio in #104
• Custom summarization prompt by @luarss in #78
• New support for 100+ model types and vendors with LiteLLM by @nkanu17 in #107

New Contributors

• @raphaeldelio made their first contribution in #104
• @luarss made their first contribution in #78
• @nkanu17 made their first contribution in #107

Full Changelog: server/v0.12.5...server/v0.12.6

0.12.5

What's Changed

Enhancements

• Add long_term_memory_strategy support to client and MCP by @abrookins in #82
• Store working memory as JSON instead of string data by @fcenedes in #95
• Allow token add to accept pre-generated tokens by @abrookins in #88
• Add JSON output for token CLI token commands by @abrookins in #86
• Support AWS Bedrock Embedding Models and LLM Models by @f4roukb in #94
• Add GHA workflow to publish server to PyPI by @abrookins in #98

Docs

• Document Docker image locations and default to development mode by @abrookins in #77
• Update Memory types link to a working page. by @allierays in #84
• Document token CLI JSON output and CI/Terraform bootstrapping by @abrookins in #93

Fixes

• Fix CLI token show command to exit with code 1 on errors in JSON mode by @Copilot in #90
• Fix MCP search_long_term_memory missing background_tasks parameter by @abrookins in #99

New Contributors

• @allierays made their first contribution in #84
• @f4roukb made their first contribution in #94
• @fcenedes made their first contribution in #95

Full Changelog: server/v0.12.3...server/v0.12.5

Docker images published:

redislabs/agent-memory-server:0.12.5
ghcr.io/redis/agent-memory-server:0.12.5
redislabs/agent-memory-server:latest\n - ghcr.io/redis/agent-memory-server:latest

0.12.3

What's Changed

• Fix background task execution in --no-worker mode by using HybridBackgroundTasks directly as type annotation instead of Depends(get_background_tasks)
• Add deduplicate parameter to create_long_term_memory endpoint with default true
• Pass background_tasks parameter through search_long_term_memory and memory_prompt endpoints
• Update MCP server to create HybridBackgroundTasks instance for memory_prompt calls
• Update tests to patch actual functions instead of BackgroundTasks class
• Add test to verify tasks run inline in --no-worker mode

Full Changelog: server/v0.12.2...server/v0.12.3

0.12.2

What's New

• Reconstruct working memory messages from long-term storage by @abrookins in #68
• Add LangChain integration with automatic tool conversion by @abrookins in #73

Bug Fixes

• Calculating remaining context works better now by @abrookins in #68
• Client error handling by @abrookins in #72

Misc.

• Defaults change: Turn off query optimization by default by @abrookins in #68

Full Changelog: server/v0.12.1...server/v0.12.2

0.12.1

Highlights

• Simpler local dev: Add --no-worker flag to api and mcp commands. This enables a single-process mode without a Docket worker, using asyncio for background tasks.
• Smoother prompts when sessions are missing: memory_prompt now creates a temporary empty WorkingMemory if a session doesn’t exist, instead of failing.

What’s New

• --no-worker option for api and mcp CLI commands; MCP in stdio mode defaults to no-worker.
• HybridBackgroundTasks class to support dual-mode background work (Docket or FastAPI).

Documentation

• Rewrote Memory Lifecycle docs to reflect server-controlled lifecycle (Docket background tasks, best practices) and tightened language; renamed and linked Memory Integration Patterns; fixed URLs and a mkdocs.yml typo.

Upgrade Notes

• Local development: You can now run without a worker:

  agent-memory-server api --no-worker
  # or
  agent-memory-server mcp --no-worker