# RossLabs projects

> AI products, agents, and developer tools built by Tyrone Ross.

## Project write-ups

- [Agent Astronomer](https://rosslabs.ai/projects/agent-astronomer/index.md): Query your local Claude Code skill, agent, and plugin library from any conversation. Wraps the Agent Astronomer CLI as MCP tools.
- [Agent Builder](https://rosslabs.ai/projects/agent-builder/index.md): Design, evaluate, and improve agentic harnesses. Methodology playbook plus a catalog of architectures, frameworks, memory substrates, and lab patterns.
- [Agent Builder Studio](https://rosslabs.ai/projects/agent-builder-studio/index.md): One app for the agent design → run → package lifecycle — a visual canvas authors and runs a governed spec; a deterministic engine packages it. Local-first.
- [Agent DOE Engine](https://rosslabs.ai/projects/agent-doe-engine/index.md): Tests several agent settings in one planned batch and reports which changes are real, which are noise, and the best combination when goals compete.
- [Agent Rally Point](https://rosslabs.ai/projects/agent-rally-point/index.md): Coordinate coding agents across models with shared claims, decisions, handoffs, and receipts.
- [Agent Studio](https://rosslabs.ai/projects/agent-studio/index.md): Sketches small agent graphs, runs them against your own local Ollama model, and writes a transcript and brief for each run — no cloud.
- [AI Operating System Quickstart](https://rosslabs.ai/projects/ai-operating-system-quickstart/index.md): A local-first Markdown workspace for durable AI context, source preservation, decisions, and search.
- [AI Persona Lab](https://rosslabs.ai/projects/ai-persona-lab/index.md): Runs independent synthetic perspectives over a product decision, interface, plan, or codebase before release.
- [AI Upskilling Pathways](https://rosslabs.ai/projects/ai-upskill-pathways/index.md): Sequences curated AI resources into four enablement stages, one path per team function, ranked by impact so limited time goes furthest.
- [AI User Personas](https://rosslabs.ai/projects/ai-user-personas/index.md): Archived two-agent prototype whose review workflow and lessons were merged into AI Persona Lab.
- [Aida](https://rosslabs.ai/projects/aida/index.md): Find an administrative workflow worth automating and compare business choices through visible criteria and calculations.
- [API Registry](https://rosslabs.ai/projects/api-registry/index.md): Local registry of authoritative API and library docs so agents verify integrations against current sources.
- [Atomize AI: News Intelligence](https://rosslabs.ai/projects/atomize-ai/index.md): Turn AI news into searchable summaries, connected topics, and trend signals.
- [Atomize News: Intelligent Search](https://rosslabs.ai/projects/atomize-news-search/index.md): An earlier AI news search product that grouped coverage into events and connected search results to trends.
- [Blog Content Scraper](https://rosslabs.ai/projects/blog-content-scraper/index.md): Extract blog content from any website for LLM consumption. Zero API costs.
- [Bookmark](https://rosslabs.ai/projects/bookmark/index.md): Context snapshots for Claude Code. Restores session state across compactions, terminal closures, and restarts with zero manual steps.
- [Build Loop](https://rosslabs.ai/projects/build-loop/index.md): Plan a code change, verify what was built, and carry failures into another checked iteration.
- [Calm Precision UX Guide](https://rosslabs.ai/projects/calm-precision-ux-guide/index.md): Grounds every UI, writing, and deck preference a coding agent makes in cited research, not taste — one doctrine spanning web, iOS, and macOS.
- [Claude Code Debugger](https://rosslabs.ai/projects/claude-code-debugger/index.md): Debugging memory that stores incidents and extracts patterns so you never fix the same bug twice.
- [Data Intel](https://rosslabs.ai/projects/data-intel/index.md): Turn a complex research question into a cited report with a visible plan, traceable claims, and explicit checks for gaps and contradictions.
- [Diagram Intelligence](https://rosslabs.ai/projects/diagram-intelligence/index.md): Turn source evidence into a typed graph, then render diagrams for GitHub, slides, and interactive inspection from the same structure.
- [Docx Builder](https://rosslabs.ai/projects/docx-builder/index.md): Claude Code plugin that builds production Word docs — locked typography, headings, tables, TOC, and callouts — via a plan, render, check, deliver workflow.
- [Easy Terminal](https://rosslabs.ai/projects/easy-terminal/index.md): Run coding agents and shells in one native Mac workspace, then return to their conversations and files.
- [Groundwork](https://rosslabs.ai/projects/groundwork/index.md): A Claude Code plugin that plans product and design decisions — a visual picker, EARS requirements, design tokens — into a handoff-ready spec for build-loop.
- [Interface Built Right](https://rosslabs.ai/projects/interface-built-right/index.md): Routes AI-generated UI work by platform, orchestrates the build, and verifies the result against real browser data — catching what a passing test misses.
- [Local Smartz](https://rosslabs.ai/projects/local-smartz/index.md): Ports Stratagem's research patterns to run entirely on local hardware — one DeepAgent, 8 custom tools, no cloud LLM — via web UI, CLI, or a native macOS app.
- [Loop Builder](https://rosslabs.ai/projects/loop-builder/index.md): Generates declarative, host-portable specs for running a multi-step non-coding process repeatably — with a built-in quality gate and a reproducible artifact.
- [Market Intel](https://rosslabs.ai/projects/market-intel/index.md): Simple summaries of news you care about. Personalized topic tracking with AI synthesis.
- [Market Research Platform](https://rosslabs.ai/projects/market-research-platform/index.md): Local-first research workspace for ingesting source files, tracing evidence, and drafting reports without sending private material to the cloud by default.
- [Mermaid Spells](https://rosslabs.ai/projects/mermaid-spells/index.md): A cursive-letter learning app for Montessori kids 4-5 — practice in a calm underwater world with a mermaid companion. No scores, no timers, no failure states.
- [Mockup Gallery](https://rosslabs.ai/projects/mockup-gallery/index.md): Rates and tracks HTML mockups locally, then hands routed design decisions and per-route artifacts straight to your Claude Code or Codex workflow.
- [Multi-Goal](https://rosslabs.ai/projects/multi-goal/index.md): Tests every metric together in one planned DOE batch, fits real effects, and picks the best configuration when goals compete.
- [NavGator](https://rosslabs.ai/projects/navgator/index.md): Map the dependencies and runtime services a code change can affect before editing them.
- [Omniparse](https://rosslabs.ai/projects/omniparse/index.md): Universal document parser. Excel, PowerPoint, Python, PDF, and directories into LLM-ready Markdown and structured data.
- [Personal LLM Wiki](https://rosslabs.ai/projects/personal-llm-wiki/index.md): Private local-first knowledge system that turns Markdown notes into an agent-readable wiki with hybrid retrieval.
- [Pptx Builder](https://rosslabs.ai/projects/pptx-builder/index.md): Builds PowerPoint decks that keep their fonts and colors on another machine, plus a diagram-mapper that matches any source artifact to the deck's palette.
- [Product Pilot](https://rosslabs.ai/projects/product-pilot/index.md): AI copilot that generates PRDs, architecture docs, and success metrics from conversational input.
- [Prompt & Model Benchmark Lab](https://rosslabs.ai/projects/prompt-model-benchmark-lab/index.md): Compare prompts and models on the same cases, then inspect the outputs, scores, cost, and latency behind each result.
- [Prompt Builder](https://rosslabs.ai/projects/prompt-builder/index.md): Prompt Policy Engine. Classify, diagnose, rewrite, and score prompts calibrated to model tier and deployment context. Callable by agents, tools, humans.
- [Prompt Test Lab](https://rosslabs.ai/projects/prompt-test-lab/index.md): A/B test prompts across LLM models and deploy winners from the same UI.
- [PsychScribe Local](https://rosslabs.ai/projects/psychscribe/index.md): A local-first clinical-scribe concept for a solo psychiatrist — capture sessions, produce clinician-reviewed notes, strictly offline by default.
- [Pyramid Principle](https://rosslabs.ai/projects/pyramid-principle/index.md): Barbara Minto's Pyramid Principle as composable writing skills. Short-form, long-form, presentation, and audit with shared source grounding.
- [Replit Migrate](https://rosslabs.ai/projects/replit-migrate/index.md): Migrate Replit apps to web (Vercel, Cloudflare, standalone) or native iOS/macOS. Risk-ordered plans with lessons encoded from real migrations.
- [Research](https://rosslabs.ai/projects/research/index.md): Token-efficient research knowledge base with SQLite FTS5, source tier scoring, and claim verification. Persists to ~/research with project symlinks.
- [RossLabs Agent Harness](https://rosslabs.ai/projects/rosslabs-agent-harness/index.md): Model-agnostic, local-first Rust agent harness: turns any local or cloud model into a coding agent with deterministic gates, routing, and sandboxed execution.
- [RossLabs Ambient Agent](https://rosslabs.ai/projects/rosslabs-ambient-agent/index.md): Prepare the next move from authorized context while keeping external decisions with the owner.
- [RossLabs LSP](https://rosslabs.ai/projects/rosslabs-lsp/index.md): Gives coding agents compiler-backed symbol search through a vendor-neutral CLI and MCP server.
- [RossLabs Operations Center](https://rosslabs.ai/projects/rosslabs-operations-center/index.md): A local-first Rust ticket board where coding agents claim tasks from one queue, leave receipts, and park anything needing approval in one lane.
- [rosslabs-ai](https://rosslabs.ai/projects/rosslabs-ai/index.md): Personal portfolio site built with Astro and Keystatic CMS. Showcases projects, posts, and experience.
- [Secrets Vault](https://rosslabs.ai/projects/secrets-vault/index.md): Store credentials on a Mac and give local tools a narrow way to use them without pasting values into prompts.
- [Showcase](https://rosslabs.ai/projects/showcase/index.md): Deprecated predecessor to Spectra for development-time screenshots, short clips, and web walkthrough assets.
- [Slide Builder](https://rosslabs.ai/projects/slide-builder/index.md): Parse old presentations to build structure for new ones. Runs locally on Mac for privacy.
- [Speak Savvy](https://rosslabs.ai/projects/speak-savvy/index.md): Practice speaking on iPhone and get specific feedback on structure, pace, clarity, and delivery.
- [Speak Then Do](https://rosslabs.ai/projects/speak-then-do/index.md): Task extraction app that turns freeform thoughts into actionable items with AI parsing and swipe-based triage.
- [Spectra](https://rosslabs.ai/projects/spectra/index.md): Captures a running app on web, macOS, iOS, or watchOS, then turns the raw screenshots and clips into a tagged, exportable demo — no editor required.
- [Stratagem](https://rosslabs.ai/projects/stratagem/index.md): Market research agent with document processing, web scraping, SEC filings, and financial analysis. Ships a research orchestrator plus four skills.
- [Style Calibrator](https://rosslabs.ai/projects/style-calibrator/index.md): A self-paced app that turns a voice-and-style calibration pack into a clickable flow, then exports a prompt to generate your personal style profile.
- [Threat Modeler](https://rosslabs.ai/projects/threat-modeler/index.md): Produces evidence-linked threat models for agentic systems and other risk-surface changes.
- [Travel Planner](https://rosslabs.ai/projects/travel-planner/index.md): Plan family trips and summer camps in one place, then share schedules through a dedicated family calendar.
- [TruePace](https://rosslabs.ai/projects/truepace/index.md): Match focus sessions to the task, use check-ins to adjust the pace, and learn from completed work.
- [Visual Test Suite](https://rosslabs.ai/projects/visual-test-suite/index.md): URL-based visual regression testing with Playwright. Configure tests via UI, run screenshots, manage baselines.

HTML version: https://rosslabs.ai/projects/