BytePlan Data Analysis

Problem

byteplan-analysis targets workflows where BytePlan data has been retrieved but there is no structured analysis artifact. It constrains an analysis session to a fixed workspace, a reproducible task list, and standardized outputs, preventing files from being scattered across ~/.byteplan, /tmp, or an arbitrary workspace root. The skill does not render Excel, PPT, Word, web, or video reports directly; instead, it produces analysisPlan.json, analysis_report.md, and report_data.json for downstream report skills.

How It Works

Core flow:
- Workspace isolation: creates ~/.byteplan/workspaces/{topic}_{timestamp}/ using {topic}_{timestamp} and writes all files into the same directory.
- Authentication and tenant check: prefers cached token in ~/.byteplan/.env, prompts for credentials if needed, re-authenticates when expired, and displays tenant info after login.
- Task-based querying: uses an LLM to generate a task list from available models, executes model queries per task, and produces a Markdown data report.
- Follow-up recommendations: generates actionable recommendations from actual data across urgency, deeper analysis, business actions, and data collection.

Boundaries

It focuses on the data layer and analysis plan rather than document layout or visual rendering. Excel, PPT, Word, web, and video outputs require the corresponding byteplan-* report skills.