Diagnose bugs and incidents using suitable captured Runtime History before focused live capture with Lightrun MCP. Use when static analysis or logs cannot explain the failure mechanism, when a hypothesis must be tested against runtime values or control flow, or when resuming Lightrun async actions. For latency, timeout, deadline, SLO, or hang diagnosis, use lightrun-slow-execution-diagnosis when available; if unavailable, continue here with best-effort evidence. Do not use for one-off production measurements (use lightrun-ask-prod when available; otherwise report the measurement-scope gap), pull-request review, code changes, setup, or deployment.
92%
Does it follow best practices?
Run evals on this skill
Adds up to 20 points to the overall score
Low
Low-risk findings worth noting
The skill exposes the agent to untrusted, user-generated content from public third-party sources, creating a risk of indirect prompt injection. This includes browsing arbitrary URLs, reading social media posts or forum comments, and analyzing content from unknown websites.
The required workflow reads Runtime History and live captures from the Lightrun MCP server, which return application variable values and call stacks that can contain attacker-submitted strings from external requests, and the agent consumes them without first selecting a specific item.
lightrun-platform/lightrun-ai
1853b1a
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.