Skip to main content

v1.0.231

Date released: February 14, 2026

New features and enhancements​

AI​

GTM agent​

Introduced the GTM Research Agent, a powerful new tool designed to automate deep-dive company research directly from your browser. The GTM Agent works in real-time, accessing the live web to give you the most current insights available. Following are the core features:

  • Live browser integration: Utilizing a dedicated Chrome extension, the agent operates within the user's active browser session. This enables the secure retrieval of data from premium, authenticated platforms such as linkedin, crunchbase, and g2—sources typically inaccessible to standard automated scrapers.

  • Autonomous lead enrichment: The agent can process high-volume lead lists sequentially. It autonomously navigates relevant web entities to extract executive profiles, firmographic data, and market sentiment, populating internal workbooks in real-time.

  • Strategic workflow orchestration: The GTM agent is designed for "human-in-the-loop" functionality. Users may configure the system to identify ideal customer profile (icp) alignment and pause for manual verification before proceeding with downstream actions, such as personalized outreach generation.

  • Operational connectivity: Research outputs are fully compatible with enterprise crm systems, including salesforce and hubspot. This ensures that enriched intelligence is immediately available to sales representatives for informed prospecting.

To get started with the GTM agent, do the following:

  1. Install the extension: Download and install the new Chrome extension.

  2. Select your leads: Open a workbook and highlight the companies you want to research.

  3. Run the GTM agent: Watch as it autonomously gathers intelligence and populates your workspace in real-time.

Added newer Claude and GPT model support​

This update introduces support for Claude Opus 4.6, which features a 1M-token context window and industry-leading performance for agentic workflows. Also added support for the GPT-5.2 series, providing enhanced logical reasoning for complex development tasks.#### Upgrade HubSpot Agent to official HubSpot MCP server

Upgraded HubSpot Agent to official HubSpot MCP server​

HubSpot agent is updated to the Official HubSpot Model Context Protocol (MCP) Server, ensuring parity with the Atlassian MCP standards utilized by the JIRA agent. This transition ensures a robust, standardized interface for all HubSpot operations, including:

  • Core objects: Optimized handling of Contacts, Companies, Deals, and Tickets.

  • Extended functionality: Enhanced support for Engagements and Custom objects.

  • Protocol compliance: All interactions now execute via MCP-compliant tool calls, increasing reliability and data integrity.

  • Advanced reasoning and operational transparency: The agent now employs an enhanced cognitive framework to manage intricate CRM workflows.

  • Deterministic planning: The HubSpot agent generates a comprehensive execution plan now before performing any actions.

  • Process visibility: Users can now monitor the HubSpot agent's thinking steps and the specific MCP tool calls initiated, providing a clear audit trail of the logic used to address complex queries such as funnel analyses or time-based trends.

  • Computational analysis via Python and Bash: To facilitate high-level reporting, the HubSpot agent now utilizes a Bash tool for the dynamic generation of Python scripts. This enables:

    • Complex data transformation: Programmatic calculation of deal stage metrics and owner performance.

    • Custom analytics: Execution of sophisticated CRM logic to prepare datasets for advanced reporting.

  • Data visualization and portability: To improve the interpretability of CRM data, this version introduces native support for Chart Generation and Data Downloads:

    • Visual analytics: Automatic generation of Bar, Line, and Pie charts to illustrate revenue trends, deal distributions, and ticket categories.

    • Export capabilities: All processed or raw datasets can be exported directly as CSV files, ensuring seamless integration with external business intelligence tools.

Bug Fixes​

Backend​

Redshift LISTAGG character limit resolution​

Issue: The system previously utilized the LISTAGG function to retrieve distinct values for single-select UI filters. However, Amazon Redshift imposes a 65,535-character limit on LISTAGG outputs. Queries exceeding this threshold failed, preventing filter values from loading in the interface.

Resolution: The data retrieval logic has been refactored to bypass database-level string aggregation:

  • Query Update: Replaced LISTAGG with a standard SELECT DISTINCT query.

  • Backend Processing: The concatenation and :: separation are now handled within the application backend.

Impact: This change eliminates query failures for high-cardinality columns, ensuring stable and scalable filter loading regardless of the volume of distinct values.

Was this helpful?