docs(google-adk): bring ADK docs to parity with adk-redis 0.0.9

[nkanu17]

The redis.io Google ADK pages were written against adk-redis 0.0.5. Since then 0.0.7 added selectable memory backends, 0.0.8 renamed the session service, and 0.0.9 raised the managed SDK floor and added cache entry IDs. This brings all five pages to 0.0.9, which is the version currently on PyPI.

Functional breakage fixed

Two of these were not just stale wording; readers copying the samples would hit failures.

• The quick start could not reach a local memory server. 0.0.7 made redis-agent-memory (managed) the default backend, but every sample passed a localhost Agent Memory Server URL with no backend override. As written, the services silently targeted the managed backend. All service and tool samples now set backend explicitly, and the self-hosted path is called out.
• create_memory_mcp_toolset no longer exists in adk-redis. The MCP samples in redis-agent-memory.md and integration-patterns.md imported it, so they raised ImportError. Both now use ADK's native McpToolset + SseConnectionParams, matching the fitness_coach_mcp example.
• RedisWorkingMemorySessionService was renamed to RedisSessionMemoryService in 0.0.8. The old name is a deprecated alias that emits a DeprecationWarning and is removed in 0.1.0. All samples use the new name, and a note documents the alias plus the moved module path.

Config tables corrected

The documented defaults did not match the code:

Parameter	Was documented	Actual
recency_boost	False	True
extraction_strategy	None	discrete (plus a fourth value, custom)
semantic_weight / recency_weight	0.7 / 0.3	0.8 / 0.2
api_base_url, default_namespace	Required	Both have defaults

Also added the missing backend, api_key, store_id, timeout, threshold, weight, and half-life fields.

Other changes

• New "Choose a memory backend" section with a feature matrix. Recency boost, auto-summarization, extraction strategies, and MCP are self-hosted only. This is the question that prompted the update.
• Documented all six memory tools. GetMemoryTool and MemoryPromptTool were missing.
• Documented invocation-user resolution from tool_context and the client-supplied id on CreateMemoryTool (0.0.9).
• Semantic caching: added cache entry IDs, CacheEntry, and delete_by_id() targeted invalidation (0.0.9), plus the required server_url on LangCacheProviderConfig.
• Examples: nine to ten, added managed_memory_quickstart, labeled each memory example with its backend, and noted the runner per example (python main.py vs adk web .). redis_search_tools wires three tools, not four.
• Noted the redis-agent-memory>=0.2.0 floor.

Verification

Every class name, config field, default value, and example was checked against adk-redis main at 0.0.9 rather than against the changelog.

I could not get a clean local Hugo build to validate the new relref, but the same ByWeight template error reproduces on an unmodified main with Hugo 0.160.1, so it is a local version mismatch rather than anything in this change. The one new relref target (/integrate/google-adk/redis-agent-memory#choose-a-memory-backend) points at an existing page and a heading added in this PR. Worth a look at the CI build.

Known drift outside this PR

The car dealership tutorial is linked three times from these pages and still uses RedisWorkingMemorySessionService with no backend field. It lives outside this repo, so it needs a separate change from its authors.

🤖 Generated with Claude Code

---

Note

Low Risk
Documentation-only changes to integration markdown; no runtime code, with the main risk being incorrect samples or relrefs rather than production behavior.

Overview
Brings the five Google ADK / adk-redis integration pages up to adk-redis 0.0.9, fixing copy-paste samples that would fail under current defaults and expanding backend and API coverage.

Memory backends and naming: Docs now treat redis-agent-memory (Redis Cloud or self-managed Data Plane) as the default and opensource-agent-memory (deprecated Agent Memory Server) as the legacy path, with explicit backend, api_key, and store_id in samples. Working memory is renamed to session memory; samples use RedisSessionMemoryService / RedisSessionMemoryServiceConfig with a note on the 0.0.8 renames. A new Choose a memory backend section compares features (MCP, auto-summarization, recency boost, etc.) and deployment options.

Integration patterns: MCP examples drop removed create_memory_mcp_toolset in favor of ADK’s native McpToolset + SseConnectionParams, with notes that MCP is Agent Memory Server–only. REST and Runner examples use managed-backend config; comparison tables add a Backends column.

Reference accuracy: Session and long-term memory config tables are expanded and defaults corrected (recency_boost, weights, extraction_strategy, optional api_base_url / namespace). REST docs list six tools (GetMemoryTool, MemoryPromptTool) plus invocation-scoped users and idempotent CreateMemoryTool IDs (0.0.9).

Examples and caching: Examples count goes to ten (managed_memory_quickstart), each labeled by backend and run mode; search example reflects three tools. Semantic caching adds CacheEntry, delete_by_id(), and required LangCache server_url.

^{Reviewed by Cursor Bugbot for commit 51c7df9. Bugbot is set up for automated code reviews on this repo. Configure here.}

[CLAassistant]

Thank you for your submission! We really appreciate it. Like many open source projects, we ask that you sign our Contributor License Agreement before we can accept your contribution.
_{You have signed the CLA already but the status is still pending? Let us recheck it.}

[mich-elle-luna]

Hi @nkanu17 thank you for this update, last week @n-valchev documented the self-hosted installation of agent memory here, https://redis.io/docs/latest/develop/ai/context-engine/agent-memory/self-managed/

Please update this PR to refer to those new pages instead of the deprecated agent server pages and repo (or in addition to that). Also just fyi, @raphaeldelio is thinking of moving all that self-managed content Nikolay published last week to the /operate section, so you all should coordinate potentially if he does to ensure that link targets are updated. Thank you!

[mich-elle-luna]

Looking good, just need to add references to https://redis.io/docs/latest/develop/ai/context-engine/agent-memory/self-managed/ and related content.

[nkanu17]

Thanks @mich-elle-luna, updated in 51c7df9.

Acting on this surfaced something bigger than a missing link, so flagging it explicitly for you and @n-valchev.

These pages presented a binary: managed means redis-agent-memory, self-hosted means opensource-agent-memory. That framing is wrong, and combined with the deprecation label it would have actively steered anyone wanting to run Agent Memory themselves toward the deprecated backend.

Self-managed Agent Memory serves the same shared Data Plane API as Redis Cloud (/v1/stores/{storeId}/..., per the self-managed API examples), and adk-redis builds its client as _AgentMemory(api_base_url, api_key=..., store_id=...) with the base URL as a positional argument, so it is not Redis Cloud specific. That means three deployment paths across two backend values:

Deployment	backend	api_base_url
Redis Cloud	redis-agent-memory	Redis Cloud Agent Memory endpoint
Self-managed	redis-agent-memory	Your own Data Plane URL
Agent Memory Server (deprecated)	opensource-agent-memory	Your server URL

You pick a deployment with api_base_url, not with backend. So the pages now:

• Add a deployment table mapping each path to its backend, api_base_url, and setup guide, linking Nikolay's self-managed pages and the Redis Cloud create-service page.
• Add a note warning against reaching for opensource-agent-memory just because a deployment is self-hosted. The two are different systems, not two deployments of one system.
• Change the feature matrix columns from "Managed vs Self-hosted" to the two backends, since the differences follow the backend. Self-managed has the same feature set as Redis Cloud.
• Retire "self-hosted" as a synonym for opensource-agent-memory across all five pages, because self-managed Agent Memory is also self-hosted.

On the /operate coordination with @raphaeldelio: I used relref rather than absolute URLs throughout, so if the self-managed content moves, the Hugo build fails loudly instead of leaving dead links. Verified all 20 relref targets on these pages resolve.

One thing worth a second opinion: the feature matrix still shows recency-boosted search, auto-summarization, extraction strategies, and MCP as Agent Memory Server only. I framed that as current state plus a migration-timing consideration, but if there is a roadmap or support timeline for those on the Data Plane API, that wording should say so rather than implying the gap is permanent.

[mich-elle-luna]

Hi Nitin, thank you for these updates. Unfortunately, I think this is really confusing with the old deprecated content still weaved into the new info. So, my feedback is to separate or keep all the old info in a section or file that is marked deprecated with a banner, and provide only the info for the supported services that will be maintained going forward in these main files (removing or moving content that really no longer applies. But, I can be persuaded otherwise , if you have perspectives from others that this content must live side-by-side. Thanks!

[mich-elle-luna]

Suggested change

	- **Memory backends** power the session and long-term memory services. Pick one per service with a `backend` field:
	[Redis Agent Memory]({{< relref "/develop/ai/context-engine/agent-memory" >}}) handles working memory (sessions), long-term memory (extracted facts), auto-summarization, and memory search.

[mich-elle-luna]

Suggested change

- **[Redis Agent Memory](https://redis.io/agent-memory/)** (`redis-agent-memory`, the default) is the Agent Memory service. Use this for new work. It runs either on [Redis Cloud]({{< relref "/operate/rc/context-engine/agent-memory" >}}) or [self-managed]({{< relref "/develop/ai/context-engine/agent-memory/self-managed" >}}) on your own Kubernetes cluster; both share one Data Plane API, so you pick a deployment by pointing `api_base_url` at the right endpoint.

Use the default `redis-agent-memory` for new work. It runs either on [Redis Cloud]({{< relref "/operate/rc/context-engine/agent-memory" >}}) or [self-managed]({{< relref "/develop/ai/context-engine/agent-memory/self-managed" >}}) on your own Kubernetes cluster; both share one Data Plane API, so you pick a deployment by pointing `api_base_url` at the right endpoint.

[mich-elle-luna]

Suggested change

	- **[Agent Memory Server](https://github.com/redis/agent-memory-server)** (`opensource-agent-memory`) is the open source memory server, now deprecated. It is documented for existing deployments and currently remains the only backend offering auto-summarization, extraction strategies, recency-boosted search, and an MCP endpoint.

[mich-elle-luna]

{{< note >}}
Agent Memory Server** (opensource-agent-memory) is now deprecated.
{{< /note >}}

[mich-elle-luna]

Suggested change

See [Redis Agent Memory]({{< relref "/integrate/google-adk/redis-agent-memory" >}}) for the feature-by-feature comparison of the two memory backends.

[mich-elle-luna]

move or remove

[mich-elle-luna]

move or remove

[mich-elle-luna]

move or remove

[mich-elle-luna]

move or remove

[mich-elle-luna]

move or remove