Basketball Shot Analyzer
A manual-plus-vision tool for tracking basketball shots from video and calculating release speed, angle, and trajectory.
These are not talent-ready showcase pages yet. Each note points to public material and still needs real artifacts, visitor-facing case-study copy, architecture, stack rationale, and next-step links before it can enter discovery.
A manual-plus-vision tool for tracking basketball shots from video and calculating release speed, angle, and trajectory.
A basketball video analysis workflow using detection, tracking, player tagging, event labeling, verification, and review-package export.
A repeatable incident workflow for diagnosing and repairing healthcare ATS source failures and production aggregation issues.
Source-level aggregation health reporting that surfaces run status, imports, worker class, and operational issues.
A bounded agent-assisted workflow for creating and validating new ATS adapters such as Hireology, e3Applicants, and Taleo.
A multi-lane autonomous operating loop for supply, social, video, SEO, product research, product build, and delivery workflows.
A crawl-depth and indexing workflow that improves discovery of live job pages and plans Google Indexing API submissions.
A production cleanup workflow for healthcare job classification, U.S. location normalization, and conservative backfills.
A programmatic video workflow for healthcare job and location content using Remotion, Sora-style assets, overlays, and rendered clips.
Internal operational tooling for health checks, reporting, Linear issue sync, and admin status views.
A public comps dashboard for HR tech, staffing, marketplaces, and robotics with weekly snapshots and disclosure-backed insight feeds.
A research system for collecting investor materials, extracting evidence, and building disclosure-backed operating metrics.
A historical backfill workflow for discovering and enriching investor-material disclosures over time.
A monitoring workflow for validating company investor-relations pages, financial pages, news pages, redirects, and freshness.
A local research inbox and skill-driven workflow for turning AI/autonomy/work source material into summaries and channel-ready drafts.
Project-scoped Claude skills for lens-driven summaries and channel-specific writing with voice memory and approval guardrails.
An agent-assisted investing research pack that moves from global universe building to ranked memos and position-ready examples.
Reusable investing analysis skills covering Mauboussin expectations, Munger inversion, Ackman quality, Seven Powers, factor quant, and more.
A labor-market analysis workspace for Appcast cost-per-application data by role, region, and healthcare specialty.
A Claude Code skill for used-item selling workflows: pricing research, listing generation, posting, message checks, and reply drafts.
An AI-assisted software delivery workflow spanning code changes, PRs, CI repair, deployment, logs, and production verification.
A collection of browser-driven research, QA, scraping, and operational verification workflows across local and production systems.
A public college-search product example held out of buyer discovery until AI-specific implementation proof is available.
A workflow from Browserbase that runs a browser agent against a real task until it converges, then graduates the winning approach into a durable markdown skill any future agent can load and run.
A Bash-based workflow that repeatedly runs Claude Code against bounded changes, pushes branches, opens PRs, waits for checks and reviews, merges green work, and carries context through a shared notes file.
A corporate engineering workflow for using Claude Code as an agentic coding workspace with planning, context management, verification, and parallel session practices.
A hosted AI product-management workflow surface for drafting and refining product artifacts.
A public workflow example covering multiple agents for CRM, outreach, podcast preparation, family scheduling, and progressive trust practices.
Railway exposes deployment context to AI coding tools through a remote MCP server and a CLI command, so agents can inspect services, logs, and service details while helping with app delivery.
A Sentry MCP workflow that gives AI coding tools access to issue, trace, and session context so agents can reason about production errors with observability data in view.
A docs-backed example library for building multi-agent crews and deterministic flows, with concrete automations for marketing, recruiting, content creation, email response, lead scoring, and meeting assistance.
An emerging maintainer workflow for using multiple agentic systems to inspect a high-volume Claude Code component repository and learn from incoming PRs.
A team-scale AI coding workflow for orchestrating multiple Claude Code sessions across worktrees and complex codebases with context-engineering practices.
A research-plan-implement workflow for shipping low-effort Linear tickets through Claude Code and GitHub Actions.
A prospective workflow builder that uses Claude Code to help create n8n AI agents and automations from natural language.
An open-source, role-based delivery loop for AI coding agents: product framing, plan review, implementation, review, browser QA, shipping, and retrospectives.
A schema-first workflow for extracting validated structured data from LLMs using Pydantic models, retries, streaming, and provider-specific integrations.
A multi-agent workflow example that partitions agents by role, device, and access boundary so each agent handles a narrow household, education, finance, coding, or operations context.
A curriculum workflow example where photos of books, supplies, and teaching materials are turned into structured lesson plans and next-step recommendations.
A screen-recording workflow example where a Loom capture of a lesson is converted into a text-first learning log for an agent to summarize and track.
A household operations workflow example signaled by a public X post about OpenClaw using Amazon, with sensitive purchasing and account details intentionally omitted.
A deployed college-search product example. It appears promising as an inspectable public app, but needs reviewer confirmation of scope, ownership, screenshots, and user value before curated Atlas case-study publication.
A docs-backed workflow system for designing durable agent loops, explicit graph state, human review points, tool execution, memory, streaming, and production observability.
A docs-backed workflow system for building event-driven agents, RAG flows, extraction flows, multi-agent handoffs, streaming events, and inspectable state over private or application-specific data.
A reusable Claude Code Skill that turns senior engineering maintenance practices into an executable workflow for AI-built projects.
A workflow tool that turns GitHub issues or coding tasks into researched requirements, confirmed subtasks, cloud-executed Claude Code changes, stacked diffs, review comments, and PRs.
A local Kanban interface that orchestrates AI coding agents through read-only proposal, read-only planning, write-enabled execution, branch isolation, diffs, testing, budgets, and autonomous task selection.
A workflow layer for AI coding agents that defines reusable steps, prompt templates, plugins, exit criteria, event feeds, resumable sessions, and e2e test generation.
A text-first workflow pattern where personal processes, examples, notes, and Markdown files are structured so an AI agent can synthesize, reorganize, and act on them over time.
A docs-backed workflow pattern for building agents with typed dependencies, function tools, structured outputs, model settings, observability, and eval-friendly boundaries.
A multi-agent development workflow that pulls Linear issues, assigns Claude Code workers/reviewers/testers/documenters, records memory, reports through Discord, and iterates on PRs.