1k-sentry-analysis_skill

This skill analyzes Sentry crash reports, identifies root causes, and prepares evidence-based bug analysis logs before requesting user confirmation.

TypeScript

2.3k

GitHub Stars

1

Bundled Files

2 months ago

Catalog Refreshed

4 months ago

First Indexed

Readme & install

Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.

Installation

Preview and clipboard use veilstrat where the catalogue uses aiagentskills.

npx veilstrat add skill onekeyhq/app-monorepo --skill 1k-sentry-analysis

SKILL.md3.8 KB

Overview

This skill analyzes Sentry crash reports and produces evidence-based bug analysis logs, then guides safe fixes for AppHang, ANR, crash, and exception reports. It focuses on reproducible root-cause identification, a required pre-change analysis artifact, and a gated workflow that waits for developer approval before code edits. The goal is reliable, testable fixes for mobile and desktop crypto wallet codebases.

How this skill works

Ingest a Sentry JSON log and extract stack traces, breadcrumbs, device metadata, and timestamps. Produce a complete bug analysis log in node_modules/.cache/bugs/ that contains six types of proof (stack trace, breadcrumbs, code logic, timing, device, and fix verification). After user confirmation, recommend or implement fixes using common patterns (concurrency control, main-thread offload, error boundaries) and provide testing and PR guidance.

When to use it

Investigating iOS AppHang events (UI frozen >5s)
Resolving Android ANR reports tied to main-thread work or deadlocks
Analyzing crashes with native or JavaScript stack traces
Gathering evidence from breadcrumbs and device metrics before fixes
Preparing an auditable bug analysis log required for production changes

Best practices

Always create a complete bug analysis log in node_modules/.cache/bugs/ before coding
Do not start code changes until explicit user or maintainer confirmation is given
Use the six-proof methodology to tie error to root cause with measurable evidence
Prefer non-invasive fixes first: limit concurrency, offload heavy work, add error boundaries
Include regression tests or manual verification steps and document verification results

Example use cases

Limit concurrent network requests when AppHang reports show many simultaneous calls
Offload heavy initialization work with InteractionManager.runAfterInteractions in React Native to fix ANRs
Add ErrorBoundary wrappers around unstable components to catch UI exceptions
Investigate a native crash by correlating device model and OS version from Sentry metadata
Generate a reproducible test case and verification checklist before creating a PR

FAQ

Yes. A complete bug analysis log saved under node_modules/.cache/bugs/ is required before any code changes.

When can fixes be implemented?

Only after the analysis log is reviewed and explicit user or maintainer confirmation is received.

What evidence should the analysis include?

Include stack trace, breadcrumbs, code logic explanation, timing/frequency data, affected devices/platforms, and a planned verification method.