Skip to content

SLayer, a semantic layer maintained by your agent

v0.9.5 Feature

This release adds 3 notable features for engineering teams evaluating rollout.

Published 19d AI Agents & Assistants
✓ No known CVEs patched
Read the diff → Tool health → What is this tool? →

✓ No known CVEs patched in this version

Topics

semantic-layer

Affected surfaces

rbac

Summary

AI summary

Row-level security now works across joins with explicit join paths.

Full changelog

SLayer 0.9.5

A small release extending row-level security and sharpening agent guidance.

  • Row-level security now works across joins: when the tenant column lives on only one table, reach it from others with a JoinFilterRule that names the join path. Cardinality-safe EXISTS semi-join, so LEFT JOIN results are never dropped or duplicated.
  • Join paths are explicit (the hops you write, not auto-discovered), and any policy with a join rule must include a blocking column rule so forgotten tables fail closed.
  • ClickHouse users: correlated subqueries need server 25.4+; SLayer probes the version and errors clearly on older servers.
  • Agents driven via MCP are now told to call help() first, following a help, search, inspect, query flow.
Additional technical details

create_mcp_server attaches its query engine as mcp._slayer_engine so long-lived hosts can await engine.aclose() at teardown to release per-task database pools. validate_models and recommend_root_model reuse this shared engine, and the schema-drift probe caches its SQL client back into the shared pool for disposal.

View diff on GitHub

Weekly OSS security release digest.

The CVE patches and breaking changes that affected production tools this week. One email, every Sunday.

No spam, unsubscribe anytime.

Share this release

Share on X Share on Bluesky

Track SLayer, a semantic layer maintained by your agent

Get notified when new releases ship.

Sign up free

About SLayer, a semantic layer maintained by your agent

All releases →

Related context

Earlier breaking changes

  • v0.7.1 Changes `search()` response to a single flat `results` list capped by `max_results`, removing separate `memories`, `example_queries`, and `entities` buckets.
  • v0.7.1 Changes `search()` to return a single flat `results` list, removing per‑bucket caps.
  • v0.6.3 Datasource names now reject dots, slashes, nulls, empty/whitespace; existing names containing '.' will fail validation on upgrade.
  • v0.6.0 recall_memories surface entirely removed with no deprecation shim.
  • v0.5.1 Two-mode reference semantics enforced: SQL mode accepts arbitrary SQL; DSL mode strictly resolves identifiers.

Beta — feedback welcome: [email protected]