Agent Interaction Trajectory Capture

via “agent-interaction-trajectory-capture”

Realistic web environment for autonomous agent testing.

Unique: Captures complete interaction trajectories (full sequences of browser actions and DOM states) rather than only final task outcomes, enabling post-hoc analysis of agent decision-making, failure modes, and behavioral patterns — supporting interpretability research beyond simple success metrics.

vs others: Richer data than binary pass/fail metrics, enabling detailed error analysis and behavioral comparison, but requires substantial storage and analysis infrastructure compared to outcome-only evaluation.

via “multi-agent-interaction-tracing”

Observability platform for AI agent debugging.

Unique: Captures inter-agent communication and coordination at the SDK instrumentation level, enabling visualization of the full execution graph of multi-agent systems without requiring agents to implement custom logging.

vs others: Provides built-in multi-agent tracing within the observability platform, whereas most multi-agent frameworks require manual logging or external tracing infrastructure to visualize agent interactions.

via “trajectory recording and agent execution tracing with hud visualization”

Open-source infrastructure for Computer-Use Agents. Sandboxes, SDKs, and benchmarks to train and evaluate AI agents that can control full desktops (macOS, Linux, Windows).

Unique: Implements a trajectory recording system that captures complete execution context (screenshots, action commands, VLM reasoning, timestamps, environment state) with HUD integration for visual overlay of agent actions on screenshots. Supports multiple export formats for compatibility with OSWorld and other benchmarking frameworks.

vs others: More comprehensive than simple logging because it captures visual context and enables deterministic replay; HUD visualization provides better debugging UX than text-only logs, while trajectory export enables standardized benchmarking vs. proprietary evaluation formats.

via “browser-interaction-recording-with-dom-state-capture”

🌐Web Agent Protocol (WAP) - Record and replay user interactions in the browser with MCP support

Unique: Captures full DOM state alongside interaction metadata at each step, enabling agents to understand both the action taken and the resulting page state — most record-replay tools only store action sequences without semantic context

vs others: Provides richer training signal than simple action logs because agents can learn from DOM deltas and element state changes, not just coordinate-based clicks

via “trajectory recording and replay for debugging and evaluation”

** - MCP server for the Computer-Use Agent (CUA), allowing you to run CUA through Claude Desktop or other MCP clients.

Unique: Implements trajectory recording as a built-in feature with support for replay, export to multiple formats, and integration with evaluation benchmarks (OSWorld), enabling systematic agent analysis and dataset creation.

vs others: More comprehensive than manual logging because it captures complete execution state; more useful than video-only recording because it includes structured data (actions, reasoning, errors) enabling programmatic analysis.

via “trajectory-based execution recording and analysis”

Library/framework for building language agents

Unique: Captures full execution context at each node including prompts, tool selections, and intermediate outputs, enabling node-level loss evaluation and targeted symbolic updates rather than only final-output feedback

vs others: More comprehensive than simple logging by structuring trajectories for analysis; enables fine-grained optimization impossible with only final-output metrics

via “agent-to-agent interaction and collision resolution”

A multi-agent environment simulation library

Unique: Uses a pluggable interaction handler pattern where collision resolution logic is decoupled from detection, allowing different interaction rules to be applied to the same collision pair based on agent types or simulation context

vs others: More flexible than physics engines like Rapier because interaction outcomes are fully customizable (agents can merge, exchange state, or trigger behaviors) rather than being constrained to physical realism