Version: 2.0

Documentation Changelog

This changelog tracks changes to Vectara's documentation — new and updated guides, context-engineering material, and API reference updates. For product and feature releases, see the Release Notes.

July 2026Latest

New Documentation

#Added pipeline source documentation for Box, Confluence, Web, and Wolken KB, refreshed the S3, SharePoint, and Google Drive source pages, and updated the pipeline sources index and session metadata page.
#Documented the slack_get_file tool and bot-token-only credentials (the slack_bot_token agent secret) on the Slack Connectors page.

June 2026

New Documentation

#Added Build with io documentation, introducing the io assistant in the Vectara Console as a conversational way to create and edit platform resources (such as agents, pipelines, and API keys), inspect your account, debug performance, and explore how Vectara works, with the confirmation flow and how to review the generated draft before saving and testing the agent.
#Added Agent Aliases documentation: a concept page covering the routed policy model (ordered rules, single and weighted targets, partition_by hashing, resolution fixed at session creation, and the independent alias and agent key namespaces) and a how-to covering direct aliases, canary rollouts, tenant routing, and policy replacement via /v2/agent_aliases
#Added Call other tools from lambda tools documentation covering the in-sandbox tool module, tool-owned versus reference tool configurations, $ref resolution against the hosting agent's secrets and session metadata, and the privacy of composed tools (no agent-surface exposure, no session events)
#Added Reusable tool configurations documentation covering the /v2/tool_configurations endpoints and attaching a shared configuration to an agent with a reference entry in its tool_configurations map
#Added Visual data documentation for ingesting image-heavy content, with links to the related chunking, tables, and data-ingestion build guides
#Added SharePoint pipeline source documentation covering Azure AD OAuth configuration and site scoping
#Added Google Drive pipeline source documentation and a pipeline session metadata page
#Documented the glossary_put_entries and glossary_delete_entries agent tools on the Glossaries page
#Documented PATCH support on the web_get tool page
#Added Enable SSO with OpenID Connect (OIDC) documentation for the Vectara SaaS deployment, covering registering an OIDC application with any provider (Okta, Microsoft Entra ID, Google Workspace, Ping Identity, OneLogin, and Keycloak), the parameters you exchange with Vectara, the callback URL, email-domain routing, group-to-access-level mapping, the permission model, and connection verification, alongside a new SaaS deployment guides index
#Documented tool server authentication and credential wiring on the Tool Servers page, covering Bearer token, custom header, and OAuth 2.0 client-credentials authentication, service-account (shared) versus per-user (session-scoped) credentials, and synchronizing, updating, and deleting tool servers

Documentation Updates

#FixedCorrected retrieval access-control claims and contrasted the ingest path with the agent-tools path
#UpdatedDocumented how to quote filter field names that contain special characters
#RemovedRemoved the fabricated "SmartChunk" term from the platform-architecture pages
#FixedFixed the Slack connector create example by dropping the top-level type field

May 2026

New Documentation

#Added a standalone UserFn language reference covering the shared expression language used by UDF rerankers, agent step transitions, and pipeline verification. Adds documentation for extract_datetime_part and the knee(sensitivity, early_bias) form, clarifies that seconds/minutes/hours are bidirectional, calls out the lack of then/ternary syntax, and summarizes each caller's context schema in one place
#Added Agent Secrets documentation covering the /v2/agents/{agent_key}/secrets resource, the {"$ref": "agent.secrets.<name>"} pattern in argument_override, masking on read and in tool_input events, and migrating credentials off agent.metadata
#Added Slack Connectors documentation
#Added chunking strategies guide and related docs links

Documentation Updates

#UpdatedDocumented fire_every and skip_first on templated reminders, including the late-conversation reminding pattern for combating context rot and the post-compaction counter reset
#UpdatedAdded a Tool Output Transform section covering jq projection of tool responses, with real web_search / web_get shapes and a search → fetch pairing example
#UpdatedPromoted tool-output gate pattern in agents/steps docs
#UpdatedAdded 'The platform' section explaining architecture, agents, and how to build
#UpdatedAdded integration-points concept doc and web_get how-to
#ChangedMoved Integrations documentation under Agents and cross-linked
#Documentation updates for agents, security, and evaluation
#ClarificationRefined onboarding pages and removed edit page links

April 2026

New Documentation

#Added Glossaries documentation covering term-to-expansion mappings, matching behavior, and attaching glossaries to agent steps via glossary_expansion reminders
#Added Pipelines documentation (quickstart, concepts, sources, verification, dead letters) for automated bulk processing of external data through agents
#Added Multi-Client Steering documentation covering multi-device session consumption, behavior: steer vs follow_up, and mid-turn interrupts
#Added Agent Skills for coding agents docs page
#Added Python SDK Agents documentation
#Added a dedicated Context Engineering section with an opinionated overview and deep-dive pages for Compaction, Reminders, Skills, Tool-Output Offloading, and Steps
#Overview covers the pillars of a well-engineered agent: retrieval, instructions, tool configuration, sub-agent delegation, long-session management, tool-output protection, and monitoring

API Reference Updates

#Updated the API Reference spec with all features released January–April 2026
#Replaced x-supported-api-key-type annotations with x-required-roles on all 133 endpoints, showing exactly which API roles, corpus roles, and agent roles grant access
#Each endpoint page now shows role badges in the authorization section

Documentation Updates

#UpdatedReorganized API Reference sidebar into grouped sections (Agents, Data, Search and Retrieval, Access Control, Analytics, Platform, Quality and Evaluation)
#UpdatedRewrote Role-Based Access Control documentation to cover all four role tiers: API roles (13 account-level roles), corpus roles (viewer/editor/administrator/owner), agent roles (agent_user/agent_viewer/agent_developer/agent_administrator), and platform roles
#UpdatedAdded API conventions and authentication documentation to the API Reference introduction page
#UpdatedAdded retrieval tuning guides and cleaned up search-and-retrieval docs
#UpdatedAdded llms.txt support for docs site
#UpdatedAdded structured output usage guide

March 2026

New Documentation

#Added Context Compaction documentation covering automatic context summarization, session forking, and token limit management
#Added Tool Versioning and Metadata documentation covering category, lineage, version, tool group metadata, and experimental tool marking
#Added Tool-Output Offloading documentation covering configurable thresholds, session artifacts, and artifact_read / artifact_grep / artifact_jq tools

February 2026

New Documentation

#Added Agent Reminders documentation covering re-injection of instructions into user inputs and tool outputs for long-session context management
#Added Multi-Step Agent Workflows documentation covering named steps, step-specific instructions, tools, reminders, and configurable transition conditions
#Added Agent Schedules documentation covering scheduled agent execution at specified intervals, session creation, and execution history
#Added Agent Skills documentation covering named skills for explicit capability invocation and deterministic tool control
#Added Concurrent Agent Sessions documentation covering multi-client interaction, interrupt handling, and collaborative workflows

January 2026

Documentation Updates

#EnhancedEnhance Agents content with updated quickstart, concepts, instructions, and tools