Fetching from the wire…
01
02
03
04
05
06
07
08
OSS2026-09-22 · source-backed
Released September 22 under Apache 2.0, fusing Elasticsearch-style search with ClickHouse-style analytics, pitched at fleets of agents each firing hundreds of queries a second rather than at human dashboard users, with roles and permissions enforced for agents at the data layer and indexing of data where it lives instead of copy-first ETL. CTO Andrey Abramov claims a design-partner workload that cost about $2,000 a month now runs at roughly a tenth of that, with configurations published on GitHub. 780 stars as of September 22. Vendor-run benchmarks until someone reproduces them. (ACCESS Newswire)
Each link below shares sources, entities, or timing with this story.
TIN reached GA v1.0.2 for all Postgres and Neki databases on September 16 and hit the HN front page at 218 points on September 19 (PlanetScale). Vendor-run benchmarks: 25x throughput and 26x lower p99 against ParadeDB on mixed queries, 36x against pg_textsearch on disjunctions...
Bounded time ranges, recursive trace/span/score predicates, thread grouping and deterministic cursor pagination, validated and rejected before any storage adapter runs, with implementations in ClickHouse, DuckDB and Postgres. Tenant-scoped trace deletion handles up to 1,000 tr...
Factory shipped into beta September 8 and took Product Hunt's top slot September 9 with 468 upvotes. The post gives audited counts: agents authored 17.0% of merged PRs (277 of 1,627) and closed 28.5% of issues (222 of 778), against a headline claim of 25-35% and 50-60%. The ch...
MemPalace (57,821 stars, v3.6.0) reports 96.6% raw recall@5 on LongMemEval with no LLM required, 98.4% with hybrid v4 on a held-out 450 questions, LoCoMo R@10 rising 60.3% → 88.9%, ConvoMem 92.9%, MemBench 80.3%, while explicitly refusing head-to-head comparison against Mem0,...
@mastra/memory gains async beforeObservation, afterObservation, beforeReflection and afterReflection hooks so an app can redact or reshape messages before model calls or persistence, which is the piece anyone with PII in agent memory has been missing. Core adds idempotent batc...
Released September 4. @mastra/evals/vitest exposes runEvals as Vitest tests with expectEvals/expectEval and a per-test score reporter, so agent quality checks run in the same pipeline as unit tests. It also adds reusable sandbox template APIs so a workspace-backed coding agent...
MindPattern daily
One email a day at 7 AM. Sources and a take on every story. Unsubscribe anytime.