By bigspinai
Render personal Claude Code reports from your local session history — a practice/archetype mirror (/persona) and a token-vs-outcome ROI report (/token-roi).
Generate a personal Claude Code practice report from your local session history.
Open a sample persona report rendered from checked-in fixture data. No pipeline run, no API calls.
Chart your Claude Code token-vs-outcome ROI over time from your local session history.
You're the orchestrator. The user invoked `/persona` (installed plugin) or
Render the persona practice mirror from checked-in sample data so the user can see what `/persona` produces without running the full analysis pipeline. Use when the user asks "what does the persona report look like?", "show me a sample/demo persona report", "preview the persona report", "can I see an example?", "is there a way to test the persona report?", or wants a quick visual of the output. No subagent tagging, no metrics computation, no Anthropic API calls — just the renderer (~500ms) against pre-committed fixtures.
Render a personal Claude Code token-ROI report from your local session history. Use when the user asks "am I getting more output per token over time?", "what's my tokens-per-PR / tokens-per-committed-line trend?", "chart my Claude Code token usage vs outcomes", or runs /token-roi. Reads ~/.claude/projects/, computes weekly token/activity/engineering-outcome metrics, and renders an HTML report with inline SVG charts. All local — no API key, no network.
Uses power tools
Uses Bash, Write, or Edit tools
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
A toolkit of skills and a Claude Code plugin (bigspin) that turns your local session history into personal reports. Everything runs on your machine — no API key required, and no data ever leaves your laptop.
Two reports ship today:
| Command | What it tells you |
|---|---|
/persona | Your practice archetype — your signature moves and how you compare against a measured baseline of 4,846 real sessions from 172 Claude Code users. |
/token-roi | Your token-vs-outcome ROI — weekly trends of token usage against engineering outcomes (committed lines, PRs, agent draft lines) and your tokens-per-unit-of-work cost ratios over time. |
Both read the session transcripts Claude Code already stores in ~/.claude/projects/ and write a self-contained HTML report (plus a short inline summary) under ~/.claude/bigspin/.
Curious what
/personalooks like before running the full pipeline?/sample-persona-reportrenders a demo from checked-in sample data in ~500ms — no analysis, no session access.
Both paths produce the same artifacts in the same place — pick whichever fits your setup.
Works best from the CLI; we don't recommend the Claude desktop app for this. Open Claude Code and run, one command at a time:
/plugin marketplace add bigspinai/toolkit
/plugin install bigspin@bigspinai
/persona
/token-roi
Each command finishes by opening its HTML report in your browser automatically.
Clone the repo and ask any agentic coding tool (Claude Code, Codex, Cursor, Copilot, …) to follow the skill you want. If you're not sure how to clone, just ask your tool to do it for you, then — from the repo root — paste one of the prompts below.
For the practice mirror (/persona):
I've cloned https://github.com/bigspinai/toolkit and am running from the repo root.
Please follow `skills/persona/SKILL.md` end to end. It will analyze my local
session history in ~/.claude/projects against the project's measured baseline
corpus and produce the report. SKILL.md handles opening persona-report.html in
my browser when it's ready.
For the token-ROI report (/token-roi):
I've cloned https://github.com/bigspinai/toolkit and am running from the repo root.
Please follow `skills/token-roi/SKILL.md` end to end. It will analyze my local
session history in ~/.claude/projects and produce the token-ROI report. SKILL.md
handles opening token-roi-report.html in my browser when it's ready.
Tip: this is a smoother experience with Auto mode on, so you don't have to approve each step — but that's your call.
All artifacts land under ~/.claude/bigspin/<run-id>/ on your machine.
/persona → ~/.claude/bigspin/persona-<timestamp>/
persona-report.html — the full slide-style report (Bigspin-branded, mobile-vertical, screenshot-friendly). Opens automatically.persona-report.md — the same report as portable markdown.persona-hero.md — a tight ~10-line summary printed inline in the chat. The thing you read first.persona-hero-card.txt + persona-hero-card.plain.txt — CLI hero card, with and without ANSI color./token-roi → ~/.claude/bigspin/token-roi-<timestamp>/
token-roi-report.html — three inline-SVG charts: weekly trend (indexed, log scale), per-session outcome distributions, and tokens-per-unit-of-work cost ratios. Opens automatically.roi.csv — anonymized per-week aggregates.token-roi-hero.md — a tight inline summary./token-roi accepts two flags: --days N (lookback window, default 90) and --no-git (skip the local-git committed-line attribution).
Everything runs locally on your machine. No upload, no telemetry, no third-party request.
~/.claude/projects/ (where Claude Code already stores it) and never copied off-disk./persona additionally uses Claude Code subagents you spawn yourself — they inherit your session, so there's no separate API key and no separate vendor. /token-roi uses no subagents and no model calls at all.~/.claude/bigspin/<run-id>/.You can audit the whole thing — it's Python plus markdown skill instructions, all bundled in the plugin.
npx claudepluginhub bigspinai/toolkit --plugin bigspinComprehensive UI/UX design plugin for mobile (iOS, Android, React Native) and web applications with design systems, accessibility, and modern patterns
Multi-model consensus engine integrating OpenAI Codex CLI, Gemini CLI, and Claude CLI for collaborative code review and problem-solving.
Ultra-compressed communication mode. Cuts ~75% of tokens while keeping full technical accuracy by speaking like a caveman.
Standalone image generation plugin using Nano Banana MCP server. Generates and edits images, icons, diagrams, patterns, and visual assets via Gemini image models. No Gemini CLI dependency required.
Write feature specs, plan roadmaps, and synthesize user research faster. Keep stakeholders updated and stay ahead of the competitive landscape.
Curate auto-memory, promote learnings to CLAUDE.md and rules, extract proven patterns into reusable skills.