raindrop-add-readonly-replay

Adds per-run readonly mode to replay production traces safely via HTTP /replay endpoint.

961|60|Updated May 1, 2026
One-click install
npx skills add https://github.com/raindrop-ai/cli --skill raindrop-add-readonly-replay
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: raindrop-add-readonly-replay
Source: https://github.com/raindrop-ai/cli/tree/main/plugin/skills/add-readonly-replay
Command: npx skills add https://github.com/raindrop-ai/cli --skill raindrop-add-readonly-replay

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Replay production traces locally against a real agent by isolating side effects with a per-run Readonly mode. This enables debugging without risking prod data or user impact.

Core Features & Use Cases

  • Per-run Readonly mode that prevents destructive side effects while replaying traces
  • Scaffolds an HTTP /replay endpoint and registers the agent in ~/.raindrop/agents.json
  • Works from Workshop's Local Agent flow and via /raindrop-add-readonly-replay

Quick Start

Run the /raindrop-add-readonly-replay command to enable the readonly replay workflow for your agent.

Frequently Asked Questions about raindrop-add-readonly-replay

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I replay production traces locally without affecting live data?

Replay production traces safely by enabling a per-run Readonly mode that prevents destructive side effects while debugging against your local agent code. This isolation ensures your replay sessions carry zero risk to production data or user impact.

What is per-run readonly mode for agent trace debugging?

Per-run readonly mode is a guard that isolates side effects when replaying production traces against a real agent. It instruments side-effect points to prevent destructive operations during local debugging sessions.

How do I set up an HTTP replay endpoint for my local agent?

Create a replay server by scaffolding an HTTP /replay endpoint and registering your agent in the ~/.raindrop/agents.json registry. This allows Workshop to route replay requests to your local agent code.

Can I debug end-to-end agent traces using Workshop's Local Agent flow?

Yes, you can debug end-to-end agent traces using Workshop's Local Agent flow by wiring a per-call readonly guard and updating your package.json. Workshop reaches your local agent through the registered agents.json routing.

What are the limitations of replaying production traces in readonly mode?

Replaying production traces in readonly mode requires instrumenting all side-effect points to ensure destructive operations are caught. If side effects are not properly wired with a per-call guard, the readonly isolation may fail to prevent unintended impacts.