yuchuangu85

yuchuangu85 / llm-anr

Public

Agent-driven Android ANR evidence extraction and AI-assisted root cause analysis pipeline.

16
4
100% credibility
Found May 15, 2026 at 16 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

A pipeline that extracts and structures Android ANR crash evidence from logs into AI-analyzed markdown reports for root cause investigation.

How It Works

1
📱 Your app freezes

You notice your Android app stops responding and grab the bugreport file from the device.

2
🔍 Feed the file to the analyzer

Simply point the tool at your bugreport zip or folder, and it scans for crash clues.

3
✨ Magic organization happens

It neatly sorts the messy logs into clear sections around the exact freeze moment, ready for review.

4
🤖 Share with your AI helper

Open the generated analysis file and chat with an AI like Claude to break down what went wrong.

5
đź“ť AI builds the story step by step

The AI examines traces, events, and logs in order, filling in insights on blocks and pressures.

âś… Get fix ideas

You receive a full report with timelines, likely causes, and practical suggestions to prevent future freezes.

Sign up to see the full architecture

4 more

Sign Up Free

Star Growth

See how this repo grew from 16 to 16 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is llm-anr?

llm-anr is a Python pipeline for agent-driven Android ANR evidence extraction and AI-assisted root cause analysis. Dump in a bugreport zip or traces, and it auto-discovers ANRs, pulls high-recall slices from traces, EventLog, logcat, AnrManager, and meminfo, then spits out structured markdown ready for LLMs like Claude to analyze in phases. Solves the slog of manually hunting ANR triggers across noisy logs, delivering auditable evidence packages with conservative candidate causes.

Why is it gaining traction?

It stands out by transforming raw dumps into LLM-friendly workspaces—no more grep marathons—with phased slots for trace, EventLog, logcat, and synthesis, plus cross-source fusion that boosts hint confidence. The interactive AI agent dispatches sub-agents for CPU/memory, locks, and Binder/IO, iterating via re-probe for tighter evidence windows. Developers hook it to their coding LLMs via simple CLI like `python scripts/anr_to_ai.py bugreport.zip`, getting reports with timelines, chains, and remediation drafts.

Who should use this?

Android engineers debugging production ANRs in bugreports or traces, especially those already piping logs to LLMs for analysis. Ideal for teams chasing input timeouts or no-focus-window issues, where manual log triage eats hours. Skip if you're not integrating AI-assisted workflows or lack LLM API keys.

Verdict

Worth a spin for ANR-heavy Android debugging—solid CLI/API entrypoints, ~270 tests, bilingual docs—but at 16 stars and 1.0% credibility, it's early-stage; expect tweaks as it matures. Pair with your LLM agent for quick wins on evidence extraction and root cause pipelines.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.