--- title: "Code: Jesse Robbins" description: "Open source projects Jesse Robbins builds and contributes to, with commits, private work, agent hours and tokens." canonical: "https://jesserobbins.com/code/" jsonld: "https://jesserobbins.com/code.json" image: "https://jesserobbins.com/code/code-og.png" doc_version: "1.0" last_updated: 2026-10-31 --- # Code This is an AI-updated overview of my open source contributions. Synced Oct 31, 2026. Public commits and PRs can be checked on GitHub. Private commits, documents, review loops, agent hours and tokens are totals from local clones and agent session archives. ## Open source - [msgvault](/code/#msgvault) (Project, Contributor): Archive a lifetime of email and chat, with offline search, analytics and AI query. 482 commits (33 public, 70 fork, 379 private). The private work is in a [greenroom](/code/#greenroom), a private repo beside the public code. 19 PRs. 52 docs. 3 refine loops. 119 agent hours. 3.36B tokens. - [greenroom](/code/#greenroom) (Skill, Maintainer): Keep private design docs, drafts and review notes in a private repo right beside the public code. 133 commits (79 public, 0 fork, 54 private). The private work here is in its own greenroom. 11 PRs. 12 docs. 4 refine loops. 32.7 agent hours. 702M tokens. - [luna-doctor](/code/#luna-doctor) (Skill, Maintainer): Diagnose LunaRoute concurrency from local agent logs, and know what to fix. 125 commits (94 public, 0 fork, 31 private). The private work is in a [greenroom](/code/#greenroom), a private repo beside the public code. 8 PRs. 5 docs. 2 refine loops. 56.4 agent hours. 996M tokens. - [maclocal-api](/code/#maclocal-api) (Project, Contributor): Apple's Foundation and MLX models behind one OpenAI-compatible endpoint on your Mac. 125 commits (15 public, 25 fork, 85 private). The private work is in a [greenroom](/code/#greenroom), a private repo beside the public code. 17 PRs. 35 docs. 4 refine loops. 102 agent hours. 1.93B tokens. - [sandvault](/code/#sandvault) (Project, Contributor): Run AI agents isolated in a macOS user account and sandbox-exec. 108 commits (24 public, 57 fork, 27 private). The private work is in a [greenroom](/code/#greenroom), a private repo beside the public code. 17 PRs. 9 docs. 3 refine loops. 43.9 agent hours. 554M tokens. - [ji-renamer](/code/#renamer) (Project, Maintainer): My rewrite of the ai-renamer project. 94 commits (84 public, 10 fork, 0 private). 27 PRs. 0.3 agent hours. 600K tokens. - [photo-restore](/code/#photo-restore) (Skill, Maintainer): AI photo restoration and upscaling, with EXIF and XMP preserved and film grain kept honest. 79 commits (54 public, 0 fork, 25 private). The private work is in a [greenroom](/code/#greenroom), a private repo beside the public code. 5 docs. 1 refine loop. 36.3 agent hours. 499M tokens. - [eg4_web_monitor](/code/#eg4-web-monitor) (Project, Contributor): A local and cloud integration for EG4 and Luxpower inverters. 74 commits (0 public, 74 fork, 0 private). 5 PRs. 1 doc. 42.6 agent hours. 989M tokens. - [refine-and-polish](/code/#refine-and-polish) (Skill, Maintainer): Track a multi-iteration roborev refine loop in a private ledger, and decide when to stop. 55 commits (35 public, 0 fork, 20 private). The private work is in a [greenroom](/code/#greenroom), a private repo beside the public code. 1 refine loop. 6.4 agent hours. 213M tokens. - [whatcable](/code/#whatcable) (Project, Contributor): A macOS menu bar app that tells you what each USB-C cable plugged into your Mac can do. 22 commits (1 public, 17 fork, 4 private). The private work is in a [greenroom](/code/#greenroom), a private repo beside the public code. 1 PR. 3 docs. 4.6 agent hours. 140M tokens. - [agentsview](/code/#agentsview) (Project, Contributor): Local-first session search, analytics and token statistics for coding agents. 13 commits (2 public, 0 fork, 11 private). The private work is in a [greenroom](/code/#greenroom), a private repo beside the public code. 2 PRs. 1 doc. 1 refine loop. 12.7 agent hours. 385M tokens. - [whoops](/code/#whoops) (Project, Maintainer): Correlates Whoop heart-rate data with coding-agent activity, to detect when I am getting mad. 12 commits (12 public, 0 fork, 0 private). 2 docs. 6.8 agent hours. 25M tokens. - [skills](/code/#skills) (Skill, Maintainer): My public Claude Code skills marketplace. 9 commits (9 public, 0 fork, 0 private). 1 PR. 2.2 agent hours. 18M tokens. - [spacedock](/code/#spacedock) (Project, Contributor): Zero-dependency agent workflows with structured approval gates and adversarial review. 3 commits (0 public, 0 fork, 3 private). The private work is in a [greenroom](/code/#greenroom), a private repo beside the public code. 1 PR. 1 doc. 2.3 agent hours. 61M tokens. - [last30days-skill](/code/#last30days-skill) (Skill, Contributor): An agent skill that researches a topic across Reddit, X, YouTube, HN, Polymarket and the web. 2 commits (0 public, 2 fork, 0 private). 2 PRs. 19.4 agent hours. 269M tokens. - [a14y](/code/#a14y) (Project, Contributor): Audits how well AI agents can discover, parse and understand a website. 1 commit (1 public, 0 fork, 0 private). 3 PRs. 2.9 agent hours. 54M tokens. ## Sitemap See [sitemap.md](https://jesserobbins.com/sitemap.md) for the full list of pages on this site.