Skip to content

@SalesforceAIResearch Salesforce AI Research

All

    Repositories list

    99 repositories

    • Beagle

      Public
      :dog: Beagle is designed to be the unified platform for the broader Recursive Self-Improvement (RSI) ecosystem. Think of it as Hugging Face Trainer for agent ev…
      pythonevolutionevaluation
      pythonevolutionevaluationbenchmarksagentsllm-agentscoding-agentsagent-harnessdarwinx
      Python
      •
      Apache License 2.0
      •3•74•1•2•Updated Sep 18, 2026Sep 18, 2026

    • gift-eval

      Public
      Jupyter Notebook
      •
      Apache License 2.0
      •126•262•9•2•Updated Sep 18, 2026Sep 18, 2026

    • agentforce-adlc

      Public
      Agent Development Life Cycle — Build, deploy, test, and optimize Agentforce agents using Claude Code skills and Agent Script DSL.
      agentforcecursor-rulesclaude-skill
      agentforcecursor-rulesclaude-skill
      Python
      •
      Other
      •41•109•3•3•Updated Sep 12, 2026Sep 12, 2026

    • CRMArena

      Public
      Official Repo for CRMArena and CRMArena-Pro
      Python
      •
      Other
      •28•149•6•2•Updated Sep 10, 2026Sep 10, 2026

    • Random-Attention

      Public
      Code for the paper "Random Attention: Rethinking KV Cache Eviction for Efficient Reasoning" https://arxiv.org/abs/2609.03430
      Python
      •
      Apache License 2.0
      •6•66•1•0•Updated Sep 4, 2026Sep 4, 2026

    • tracelab

      Public
      Reference implementation, benchmarks, and traces for "Parsing the Stream: A Live Trace Model for Long-Horizon Agents and Their Observers" (arXiv:2609.01466). Ap…
      benchmarktracingarxiv
      benchmarktracingarxivllm-agentscontext-managementagent-observabilitylong-horizon-agentsagent-traces
      Python
      •
      BSD 3-Clause "New" or "Revised" License
      •0•0•0•0•Updated Sep 2, 2026Sep 2, 2026

    • SalesSim

      Public
      Python
      •
      Apache License 2.0
      •0•2•0•0•Updated Aug 28, 2026Aug 28, 2026

    • SCUBA

      Public
      SCUBA: Salesforce Computer Use Benchmark
      benchmarkcrmcomputer-use-agent
      benchmarkcrmcomputer-use-agentbrowser-use-agent
      Python
      •
      Apache License 2.0
      •3•11•3•1•Updated Aug 26, 2026Aug 26, 2026

    • self-improve-fragility

      Public
      Python
      •
      Apache License 2.0
      •0•3•0•0•Updated Aug 11, 2026Aug 11, 2026

    • claimwriter-deep-research

      Public
      Python
      •
      Other
      •0•0•0•0•Updated Aug 5, 2026Aug 5, 2026

    • AnchorBench

      Public
      Python
      •
      Other
      •0•0•0•0•Updated Jul 30, 2026Jul 30, 2026

    • oss-template

      Public template
      Apache License 2.0
      •0•4•0•0•Updated Jul 22, 2026Jul 22, 2026

    • EVQA

      Public
      Python
      •
      Other
      •0•6•0•0•Updated Jun 27, 2026Jun 27, 2026

    • DreamHouse

      Public
      Python
      •
      Apache License 2.0
      •2•1•0•0•Updated Jun 25, 2026Jun 25, 2026

    • MCP-Universe

      Public
      MCP-Universe is a comprehensive framework designed for RL training, benchmarking, and developing AI agents for general tool-use.
      Python
      •
      Apache License 2.0
      •89•600•18•15•Updated Jun 23, 2026Jun 23, 2026

    • MAS-Orchestra

      Public
      MAS-Orchestra: train, inspect, and vibe-code multi-agent systems with RL-learned orchestration and MASBench
      orchestrationmulti-agent-systemsrl-training
      orchestrationmulti-agent-systemsrl-training
      Python
      •
      Apache License 2.0
      •5•27•0•0•Updated Jun 15, 2026Jun 15, 2026

    • MTA-Agent

      Public
      Python
      •
      Other
      •1•4•1•0•Updated Jun 2, 2026Jun 2, 2026

    • CaOPD

      Public
      CaOPD: Calibration-Aware On-Policy Distillation
      agentcalibrationuncertainty-quantification
      agentcalibrationuncertainty-quantificationpost-trainingconfidence-estimationllmon-policy-distillation
      Python
      •
      Other
      •3•17•1•0•Updated Jun 2, 2026Jun 2, 2026

    • enterprise-sales-copilot

      Public
      Python
      •
      Apache License 2.0
      •2•4•0•0•Updated Jun 2, 2026Jun 2, 2026

    • vibepass

      Public
      Python
      •
      Apache License 2.0
      •1•2•0•0•Updated Jun 2, 2026Jun 2, 2026

    • enterprise-realtime-voice-agent

      Public
      Python
      •
      Apache License 2.0
      •13•62•1•0•Updated Jun 2, 2026Jun 2, 2026

    • VoiceAgentRAG

      Public
      Python
      •
      Apache License 2.0
      •30•154•0•0•Updated Jun 2, 2026Jun 2, 2026

    • gui-process-automation

      Public
      HTML
      •0•1•0•0•Updated Jun 2, 2026Jun 2, 2026

    • AudioCapBench

      Public
      Python
      •
      Apache License 2.0
      •1•0•0•0•Updated Jun 2, 2026Jun 2, 2026

    • PLATE

      Public
      Python
      •
      Apache License 2.0
      •1•7•0•0•Updated Jun 2, 2026Jun 2, 2026

    • LeastLoadedEP

      Public
      Python
      •
      Apache License 2.0
      •5•18•2•0•Updated Jun 2, 2026Jun 2, 2026

    • SCOPE-Persona-Framework

      Public
      Python
      •
      Other
      •1•6•0•2•Updated Jun 2, 2026Jun 2, 2026

    • FOFPred

      Public
      Python
      •
      Other
      •3•40•2•0•Updated Jun 2, 2026Jun 2, 2026

    • ActiveVideoPerception

      Public
      Official Code for paper "Active Video Perception: Iterative Evidence Seeking for Agentic Long Video Understanding""
      Python
      •
      Other
      •3•20•0•0•Updated Jun 2, 2026Jun 2, 2026

    • LoCoBench-Agent

      Public
      LoCoBench-Agent: An Interactive Benchmark for LLM Agents in Long-Context Software Engineering
      Python
      •
      Apache License 2.0
      •6•23•1•0•Updated Jun 2, 2026Jun 2, 2026
    ProTip! When viewing an organization's repositories, you can use the props. filter to filter by custom property.