Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    tracescope agent playbook

    1

    Automate impact analysis, Figma-to-code UI reviews, and crash triage using TraceScope MCP tools.

    Free

    1 useSecurity scanned
    tracescope agent playbook

    tracescope agent playbook

    Example session with this skill installed

    Compare the latest commits between v1.2.0 and v1.3.0 and give me a manual test checklist for the affected modules.

    • Read your context and instructions
    • Compiled the tracescope agent playbook
    • Generated the document

    I have analyzed the diff between tags v1.2.0 and v1.3.0 using tracescope_get_diff. The changes are concentrated in the Auth provider and the Navigation sidebar. I have published the full report to your TraceScope panel.

    Manual Test Checklist: v1.3.0 Impact

    ModuleComponentChange TypePriorityTest Case
    AuthuseSession.tsLogicHighVerify token refresh logic on 401 response
    AuthLoginForm.tsxUI/RefactorMediumEnsure validation errors display for malformed email
    NavigationSidebar.tsxLayoutLowCheck responsive toggle behavior on mobile breakpoints
    Sharedapi-client.jsConfigHighTest connectivity against the new production endpoint

    Analysis Summary

    • Affected Files: 4 files modified, 122 lines changed.
    • Risk Level: Medium. Changes to the core Auth provider could affect all protected routes.
    • Visual Impact: Minor changes to Sidebar padding; no Figma updates detected for these commits.

    Next steps

    1. Run tracescope_open_panel to view the interactive impact map.
    2. Trigger the Auth test suite to verify no regressions in the session handler.
    3. Check LoginForm against the v1.3 Figma design for padding consistency.

    LoginForm.tsx

    PDF · document

    Generated

    Example file from a real run - the skill writes it into your workspace.

    Connects securely to your tools. The creator never sees your data.

    About this skill

    Manual regression testing and UI audits take too long when you have to track down every changed file or design spec. This skill orchestrates TraceScope MCP to automate impact analysis, visual design reviews, and crash triage.

    What it does

    • Impact Analysis generates manual test checklists and regression scopes based on git diffs between versions.
    • Visual Audits compares Figma or Lanhu designs against code implementations to find UI discrepancies.
    • Crash Triage analyzes stack traces from Umeng or Crashlytics to identify root causes.
    • Agent Orchestration manages the execution order of TraceScope MCP tools to ensure data is published correctly.

    How it works

    1. Identify Intent by detecting keywords like impact, UI review, or crash logs in the conversation.
    2. Select Pillar to load specific workflows for Impact, Visual, or Crash analysis.
    3. Execute Tools using TraceScope MCP to pull real diffs, design snapshots, or symbolicated stacks.
    4. Publish Results via tracescope_publish_* tools to sync data with the TraceScope panel.

    Frameworks & tools

    Requires Node.js 20 or higher and the @rebornace/tracescope-mcp package. Works with Figma, Lanhu, Umeng, and Crashlytics integrations.

    Why this beats prompting it yourself

    Generic agents often hallucinate file changes or ignore design specs when asked for impact. This skill enforces a strict tool-calling sequence that uses actual git data and design tokens, ensuring checklists are based on code reality rather than guesses.

    Use cases

    • Generate a manual testing checklist for a specific commit range.
    • Compare a Figma frame against a React or Vue implementation for CSS mismatches.
    • Debug a production ANR by feeding a raw Umeng stack trace into the agent.
    • Map specific UI pages to their corresponding source files for faster onboarding.

    Known limitations

    Design comparisons are static and do not perform live browser or device screenshots. Large trees require paginated fetching via offset and limit parameters. Requires local MCP installation.

    Requires Node.js 20+ and @rebornace/tracescope-mcp. Design audits are static, not live browser/device QA. Credentials must be provided via local config or user input.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before any payment.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    1 use

    Times used to date

    Free forever

    No account required to browse

    Trust & safety

    Security scan

    Security scan passed 1 month ago

    • Free with an Agensi account

    Requires

    Node.js
    @rebornace/tracescope-mcp

    Permissions

    Read Files
    Network Access
    Listed1 month ago

    What's inside

    Frequently Asked Questions

    This skill automates the tedious parts of the QA lifecycle—specifically mapping code changes to test cases, comparing UI implementations against design tokens in Figma/Lanhu, and symbolicating crash logs for faster debugging.

    You will need Node.js 20+, the `@rebornace/tracescope-mcp` package, and a host that supports the Model Context Protocol (MCP), such as Claude Desktop or other compatible AI IDEs.

    Unlike standard prompting, this skill uses a specialized tool-calling sequence that pulls real-time Git diffs and design specs, preventing the AI from hallucinating file changes or missing CSS discrepancies.

    The skill includes the specialized agent orchestration logic, pre-defined workflows for impact, visual, and crash analysis, and instructions for integrating with TraceScope MCP tools.

    Currently, the visual audit tool performs static comparisons between code and design snapshots; it does not capture live screenshots from a running browser or mobile device.

    Yes, the skill is designed to handle large codebases and design files by utilizing paginated fetching via offset and limit parameters to maintain performance and stay within context windows.

    New in LLM Ops