AI.info
AI Tools Updates — Page 104
What changed in the AI tools this week — dated, from the makers’ own changelogs.
Auf Englisch
- Culture Amp Engage — More Relevant Manager Reviews with Conditional Follow-Up Questions
Use display rules in Performance manager reviews to show follow-up questions based on a manager’s earlier answers.
- Culture Amp Engage — Add 1-on-1 Topics in Bulk
You can now add the same topic to multiple 1-on-1 conversations in one step.
- Culture Amp Engage — Goals in Reviews Now Available
Goals in Reviews brings relevant employee goals into Performance Reviews.
- Culture Amp Engage — Custom Demographics Now Available in Performance Reviews
Use standard or custom account demographics in cycle setup and participation rules.
- Scribe — Scribe browser extension 104.31.0
New plan tiers include bundled Pins, SCIM support, and Developer Access; captured steps can be appended to an existing Scribe.
- Perceptyx AI Agent Suite — Perceptyx Launches Perceptyx Anywhere
Perceptyx launched Perceptyx Anywhere, an architecture for building AI with workforce data, evaluation capabilities, and workforce-specific models.
- Factory Droids — CLI v0.229.0, Desktop v0.186.0: Custom model management, PDF viewer, quoted replies, voice dictation key, plus terminal and app fixes
Added terminal custom model management, a PDF viewer, quoted replies, a voice dictation key, and other features.
- Honeycomb — Manage SLO Burn Alerts and Budget Resets with MCP
Agents can create and update SLO Burn Alerts and reset SLO budgets through Honeycomb MCP.
- Kapwing — Release Notes - Kapwing MCP
Kapwing MCP went live at kapwing.com/mcp, letting chatbot agents access workspaces, upload assets, create projects, and download exports.
- Manus — Introducing Manus Flex: Use your own API key in the Manus workspace
Use your own API key in the Manus workspace
- Gemini Enterprise Agent Platform — Custom denial messages for semantic governance policies
Configure a fixed Denial message on a semantic governance policy to show end users when the policy denies a request.
- Tines Workflow Automation — 🤖 Claude Sonnet 5.5 can now be selected as a model option for the default AI provider
Claude Sonnet 5.5 is now available as a model option for the default AI provider.
- Ask Sage — v3.9.6
Updated in-app Terms of Use & Privacy Policy
- H2O Driverless AI — Version 2.5.1 (September 29, 2026)
Added early stopping to CatBoost fitting; increased image fine-tuning epochs; fixed target scaling, transformer fine-tuning, imports, and date sorting.
- Distro — 🧰 Advanced Options for the Concierge embed snippet
The Concierge embed now has an Advanced Options builder that generates a snippet with configured JavaScript options.
- Agent Studio — Sierra joins the OpenAI Marketplace
Eligible customers can build agents with Sierra using their OpenAI commitments through the OpenAI Marketplace.
- UiPath Maestro — Patterns library for Maestro BPMN
Insert prebuilt, editable process structures into a Maestro BPMN process, including AI decisions with human review, approval chains, and failure handling.
- incident.io On-call — New and improved custom dashboards
New tools and features let you build more custom dashboards.
- Concierge — 🧰 Advanced Options for the Concierge embed snippet
Added an Advanced Options builder to the Concierge embed snippet that generates JavaScript options.
- incident.io Incident Response — New and improved custom dashboards
Shipped tools and features for building custom dashboards.
- Claude Code — 2.1.285
- Gemini CLI — Release v0.63.0-preview.0
- Gemini CLI — Release v0.62.0
Fixes task metadata handling, MCP tool-call titles, OAuth refresh-token retention and credential deletion, and UI border rendering.
- Guru — Connect an MCP server once, and your whole team is ready to go
Admins can share an MCP server connection with everyone using a Knowledge Agent.
- ASI:One — Make Your AI Private
Your personal AI is private until you choose to open it up. Only you and your friends can interact with it.
- Voiceflow — Transcript version filter
The Search transcripts endpoint on the Analytics API can now search across several versions in one request.
- Voiceflow — Global prompt timezone
Your agent’s timezone can now be set from the global prompt editor, right next to the prompt that uses it.
- Voiceflow — Tool execution API
You can now run a function, API tool, MCP tool or integration tool on its own, from the API, the MCP server or Atlas, and read what it returned.
- Voiceflow — Messages and prompts API
Messages and prompts can now be read, created, edited and deleted by ID through the API, the MCP server and Atlas.
- Crush — v0.97.0
Added UI controls to turn MCP servers off and display Git branches.
- OpenHands Cloud — openhands: v0.73.0
Exposed auto-disable and sandbox cleanup settings, and added ENABLE_BYOR_EXPORT as a KOTS config item.
- Pinecone — Namespace aliases
A namespace alias points at a namespace and can be atomically repointed to a different namespace in the same index, with no redeploy.
- QA Wolf — Your coding agent can now use the QA Wolf MCP
Our MCP server is live and ready to use inside the agentic coding tool of your choice.
- CodeRabbit — AI Deep Scan finding provenance | Enterprise, CodeRabbit Security
AI Deep Scan code findings can include provenance identifying the scan, primary repository revision, and selected context repositories.
- CodeRabbit — Triage for GitLab | Cloud and Self-Hosted | Team, Beta
Triage is available in beta for GitLab Cloud and self-managed organizations on the Team plan and higher.
- CodeRabbit — Triage Rules | GitHub, GitLab | Team
Triage rules can merge low-risk pull requests, close stale pull requests, and ask the Coding Agent to fix inactive pull requests.
- Vectara Platform — Browser Origin Controls for Widget Connectors
Widget connectors support configuration.allowed_origins to restrict browser origins.
- Vectara Platform — OAuth Client-Credentials Options for External LLMs
OAuth client-credentials authentication for external LLMs now accepts client_auth_method and an optional audience.
- Vectara Platform — LLM Quotas in List Responses and API Contracts
GET /v2/llms now returns each model's configured quota fields.
- beehiiv — Paid Recs, beehiiv AI, and More
Better ways to grow, earn, and get work done.