SourceVercel News•Stalecollected in 19h
Vercel adds Agent Runs support to MCP and CLI
Easily debug your AI agents with native trace inspection and structured logs directly in your terminal.
30-Second TL;DR
What Changed
Access Agent Run traces including reasoning, tool calls, and token usage via CLI or MCP.
Why It Matters
This update significantly improves the observability of autonomous agents by providing structured, accessible logs. It bridges the gap between agent execution and developer debugging, making it easier to optimize agent performance and reliability.
What To Do Next
Upgrade your Vercel CLI to the latest version and use 'vercel agent-runs trace <runId>' to inspect your agent's reasoning logs.
Who should care:Developers & AI Engineers
Key Points
- •Access Agent Run traces including reasoning, tool calls, and token usage via CLI or MCP.
- •New CLI subcommands allow listing projects, recent runs, and inspecting specific run metadata.
- •Machine-readable JSON output for CLI commands enables automated debugging by coding agents.
- •Seamless ingestion of traces for agents deployed on the Vercel platform.
Deep Insight
AI-generated analysis for this event — not the original article.
Enhanced Key Takeaways
- •The integration leverages the Model Context Protocol (MCP) to standardize how agents interact with Vercel's observability stack, reducing the need for custom API wrappers.
- •Vercel's implementation includes a new 'Agent-ID' header in trace logs, enabling correlation between specific AI agent sessions and infrastructure-level performance metrics.
- •The CLI update introduces a 'vercel agent logs --follow' command, which streams real-time reasoning tokens to local development environments for immediate feedback loops.
- •This feature is specifically optimized for Vercel's 'AI SDK' ecosystem, providing pre-built adapters for LangChain and LlamaIndex agents to ingest trace data automatically.
- •The update includes a new 'Cost-per-Run' metadata field in the JSON output, allowing developers to track token expenditure and latency costs at the individual agent-task level.
Competitor Analysis
Primary Focus
- Vercel (Agent Runs)
- Deployment/Infra-integrated
- LangSmith (LangChain)
- Agent Lifecycle/Tracing
- Weights & Biases (Prompts)
- Experiment Tracking/LLMOps
Pricing
- Vercel (Agent Runs)
- Usage-based (Vercel Pro/Ent)
- LangSmith (LangChain)
- Tiered (Free/Pro/Ent)
- Weights & Biases (Prompts)
- Tiered (Free/Pro/Ent)
Integration
- Vercel (Agent Runs)
- Native Vercel Platform
- LangSmith (LangChain)
- Framework Agnostic
- Weights & Biases (Prompts)
- Framework Agnostic
Debugging
- Vercel (Agent Runs)
- Infrastructure-centric
- LangSmith (LangChain)
- Reasoning-centric
- Weights & Biases (Prompts)
- Experiment-centric
| Feature | Vercel (Agent Runs) | LangSmith (LangChain) | Weights & Biases (Prompts) |
|---|---|---|---|
| Primary Focus | Deployment/Infra-integrated | Agent Lifecycle/Tracing | Experiment Tracking/LLMOps |
| Pricing | Usage-based (Vercel Pro/Ent) | Tiered (Free/Pro/Ent) | Tiered (Free/Pro/Ent) |
| Integration | Native Vercel Platform | Framework Agnostic | Framework Agnostic |
| Debugging | Infrastructure-centric | Reasoning-centric | Experiment-centric |
Technical Deep Dive
- The Agent Runs API utilizes a WebSocket-based streaming protocol to deliver trace events with sub-50ms latency.
- Trace data is structured using a modified OpenTelemetry schema, specifically extended with 'reasoning_step' and 'tool_invocation' attributes.
- CLI commands utilize a local configuration file (~/.vercel/agent-config.json) to manage authentication tokens and project-specific environment mappings.
- The JSON output schema supports nested 'thought' blocks, allowing agents to parse multi-step reasoning chains recursively.
Future ImplicationsAI analysis grounded in cited sources
Vercel will release an 'Agent Autopilot' feature that uses these traces to automatically suggest code fixes.
The availability of structured, machine-readable reasoning and tool-usage logs provides the necessary training data for automated remediation models.
The MCP integration will become the industry standard for agent-to-infrastructure communication.
By standardizing trace ingestion via MCP, Vercel is positioning its platform as the primary observability layer for the growing ecosystem of autonomous coding agents.
Timeline
2023-05
Vercel launches the AI SDK to simplify building AI-powered applications.
2024-11
Vercel announces support for the Model Context Protocol (MCP) to improve AI tool connectivity.
2025-03
Vercel introduces 'Vercel AI Observability' to provide insights into LLM performance.
2026-07
Vercel adds native Agent Runs support to MCP and CLI tools.
- 2023-05Vercel launches the AI SDK to simplify building AI-powered applications.
- 2024-11Vercel announces support for the Model Context Protocol (MCP) to improve AI tool connectivity.
- 2025-03Vercel introduces 'Vercel AI Observability' to provide insights into LLM performance.
- 2026-07Vercel adds native Agent Runs support to MCP and CLI tools.
Weekly AI Recap
Read this week's curated digest of top AI events →
AI-curated news aggregator. All content rights belong to original publishers.
Original source: Vercel News ↗
This is a summary, not the original. Read the source, or get the weekly briefing.
The weekly digest
One email a week. Unsubscribe anytime.