Gracker/SmartPerfetto

AI-assisted Perfetto analysis with Web UI, CLI, evidence workflows, reports, and portable runtimes.

What it solves

SmartPerfetto provides an AI-powered analysis layer for Android Perfetto traces. It simplifies the complex process of investigating performance issues like scrolling jank, slow startup, Application Not Responding (ANR) errors, and memory leaks by allowing developers to ask natural-language questions and receive evidence-backed answers.

How it works

The tool integrates with Perfetto's existing timeline and SQL workflow. It uses a system of deterministic YAML-based "Skills" and Markdown strategies to separate factual evidence (SQL results and skill outputs) from the AI model's interpretation. It can optionally integrate with local source code to provide code-aware analysis, sending relevant snippets to the AI service for root-cause reasoning.

Who it’s for

Android developers and performance engineers who need to analyze system traces to identify bottlenecks and root causes of performance regressions.

Highlights

  • AI-Assisted Investigation: Use natural language to analyze jank, ANRs, and interaction latency.
  • Evidence-Backed Conclusions: Combines SQL results and skill outputs to ensure conclusions are grounded in trace data.
  • Code-Aware Analysis: Optionally links trace occurrences to local source code for deeper investigation.
  • Multi-Interface Support: Available via a Web UI, a command-line interface (CLI), and HTTP/SSE APIs.
  • Deterministic Skill System: Uses YAML Skills to ensure consistent and factual analysis processes.

Related

  • Project
    WILLOSCAR/research-units-pipeline-skills

    A framework for agent-assisted research that transforms research goals into verifiable, resumable runs with a full evidence trail and pinned pipeline contracts.

  • Project
    chaxiu/munk-ai

    A local-first, self-improving AI testing engine that automates product validation across Android, iOS, and Web by turning natural-language intent into real-device execution.

  • Project
    kropdx/reflection-engine

    A specialized prompt that analyzes an AI assistant's memory and chat history to provide a candid, evidence-based portrait of the user's personal patterns and blind spots.

  • Project
    astrio-labs/forall

    An AI-powered formal verification tool that automates the creation of specifications and proofs to provide machine-checkable evidence of software correctness.

  • Project
    droid-ash/finalrun-agent

    An AI-driven CLI that tests Android and iOS apps using natural language specifications, utilizing multimodal models to interact with the UI and generate test reports.