blog: Persistent Memory for the Vercel AI SDK in Five Tools#2374
Merged
Conversation
Add a dedicated integration post for @vectorize-io/hindsight-ai-sdk. Covers the five memory tools (retain, recall, reflect, getMentalModel, getDocument), the semantic-vs-infrastructure input split, setup, and generateText/streamText/ToolLoopAgent/Next.js usage. Placeholder cover. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Adds a dedicated integration blog post for the Vercel AI SDK (
@vectorize-io/hindsight-ai-sdk). This was a genuine gap — the SDK is one of the largest audiences in the catalog and had zero dedicated coverage (only passing mentions in a release note and the chat post).Post:
hindsight-docs/blog/2026-06-23-vercel-ai-sdk-persistent-memory.mdSlug:
/blog/2026/06/23/vercel-ai-sdk-persistent-memoryTitle: "Persistent Memory for the Vercel AI SDK in Five Tools"
Angle
Unlike the hook-based integrations (Cline, MS Agent Framework), the AI SDK is a tool-calling framework, so memory is exposed as tools — the native idiom. The post's thesis is the integration's actual design strength: the semantic-vs-infrastructure split. The agent controls semantic inputs (what to remember / search / reflect on); the application locks infrastructure (bank ID, budget, tags, async) at construction. The model can't change the bank or blow the token budget.
Coverage
retain,recall,reflect,getMentalModel,getDocumentuvx hindsight-embed)generateText,streamText,ToolLoopAgent, and a per-user Next.js route handlerAccuracy checklist
src/tools/index.ts, the README, andexamples/integrations/ai-sdk.ts— signatures match (createHindsightTools({ client, bankId }), per-tool option groups)vitest run)npm run buildgreen, no broken links from this post (the only build break on this branch is a pre-existing strayelizaWIP file unrelated to this PR)—are empty table-cell placeholders, matching the docs page)/sdks/integrations/ai-sdkresolves; cross-links absoluteOpen items
static/img/blog/vercel-ai-sdk-persistent-memory.png— swap for final art🤖 Generated with Claude Code