Rilo LogoRilo
Get Started

Changelog

What we've shipped. Every entry corresponds to real, production-deployed work.

RSS Feed
2025-10

Security and edge-case audit rounds 1e–1j

Fix

Completed security audits for memory, observability, ingress, portal API, email gateway, and agent/LLM layers. Hardened org-boundary enforcement, secret redaction in logs and telemetry, replay protection, and resource limits.

2025-09

Security and edge-case audit rounds 1a–1d

Fix

Module audits and remediations for tools, execution engine, planner/interaction layer, and secrets/auth orchestration. Fixed request payload/header leakage, cross-tenant waiter keying, and Slack OAuth rendering sanitization.

2025-08

Tool primitive architecture and sandbox migration

Feature

Completed taxonomy and architecture deliverables, sandbox-provider migration to E2B, artifact extraction and chat delivery, objective-driven CodeExecutionTool evolution, and outbound customer-identity email send flows.

2025-07

Go-to-market onboarding pipeline

Feature

Shipped onboarding research and implementation including landing CTA canonicalization, entitlement projection and enforcement, portal UX gating, canonical mail-domain support, and email-gateway canonical routing.

2025-06

Auth strategy layer — dynamic credential forms and health lifecycle

Feature

Implemented auth strategy architecture across schema/resolver/policy layers, dynamic credential forms, challenge-interaction execution paths, and connection-health lifecycle emission and read APIs.

2025-05

Closed-loop learning and memory improvements

Improvement

Shipped closed-loop learning pipeline where execution artifacts feed back into planning, along with memory normalization improvements and heuristic cleanup across the memory subsystem.

2025-04

Platform reliability and planning stabilization

Improvement

Reliability hardening, hierarchical planning stabilization, phase elaboration, smoke coverage, and plan-validation robustness across harness and execution paths.

2025-03

Prompt and model evaluation infrastructure

Improvement

Shipped prompt/model evaluation framework, closed-loop learning and memory improvements, heuristic cleanup, email ingress security, and broad test/marker maintenance.

2025-01

Multi-tenant platform with fail-closed isolation

Feature

Launched org-isolated multi-tenancy with fail-closed credential handling, audit logging, and browser session isolation across all tool executions.