MODEL
ISSUE Nº 01 — OCTOBER 15, 2026
Three tools. Tested. The rest is noise.
Editor's letter. Hundreds of AI tools came out this week. We kept three. One we put through a fitting ourselves. The other two we couldn't try on, so they're labelled Not tested and everything about them comes from their makers.
THE COVER: Graphicle
A code graph and memory for your coding agent.
Where: https://github.com/graphicle-ai/graphicle (v0.1.0 released Oct 14, 2026; Show HN Oct 14)
The cut: One local Rust binary. It indexes your repo into a code graph and gives your coding agent (Claude Code, Codex, Cursor, VS Code, any MCP client) cited, line-anchored excerpts, so the agent stops re-reading the whole repo.
The price: Free and open source. No account, no API key, no telemetry (per README).
THE FITTING: ✅ Tested hands-on. We downloaded the v0.1.0 Linux binary (checksum verified) and ran it on a tiny 2-file Python repo:
graphicle setup --yestook 0.2s. It listed every file it touched (AGENTS.md, git hooks) and said no agent client was installed on our box.graphicle graph extract . --code-onlyproduced 6 nodes and 10 edges.graphicle ask "how are failed requests retried?"returned both files with SHA-256-tagged, cited excerpts within a 3,000-token budget. It doesn't write an answer unless you configure an LLM.
What we didn't test: the MCP integration with a live agent, and speed on a large repo. The "8x faster than Graphify on Kubernetes" claim is theirs, not ours.
The verdict: Try it. The most useful of the three for vibe coders whose agents keep getting lost in a growing codebase. You can undo it with graphicle disconnect.
Not this season if: you're on native Windows (WSL2 only, untested by them), your projects are small enough that the agent already sees everything, or you won't trust a v0.1.0 tool on your main repo. Try it on a side repo first.
THE COLLECTION
Rill Browser
A browser where Claude Code and Codex work beside you.
Where: https://rill.love (launched on Product Hunt in early Oct 2026, 500+ upvotes)
The cut: A Mac browser. Press ⌘E on any page and it becomes a task for your Claude Code or Codex. Rill picks the right project from your agent history, and the agent works in background tabs while you keep browsing.
The price: "Rill is free." AI features run on your existing Claude Code or Codex plan. Beta.
❌ Not tested hands-on. It's macOS 14+ on Apple silicon only, and we had no Mac available. Everything here comes from the official site.
How to wear it: You're reading a release note, or looking at a UI you like. Hit ⌘E, then "try this in my portfolio project," and the coding agent gets the page context with no copy-pasting. Per the site, new tasks are read-only until you allow edits, and web tasks stop before paying, posting, sending or deleting.
The verdict: Worth a download if you already pay for Claude Code or Codex. It turns "I saw something cool" into a queued agent task.
Not this season if: you're on Windows, Linux or an Intel Mac; you don't use Claude Code or Codex; or you're uneasy about an agent browsing with your sign-ins. The site itself says its prompt-injection safeguards "lower the risk; [don't] remove it."
Retriever AI (rtrvr.ai)
A free AI browser agent.
Where: https://rtrvr.ai (on Product Hunt Oct 14, 2026)
The cut: A browser agent (Chrome extension, cloud browser, API, MCP) for web chores: collecting site data, researching leads, filling forms from documents, prepping job applications.
The price: Free Mode includes supported AI models and cloud browser time with no API key or subscription, paid for by sponsored cards, with a daily fair-use allowance. Schedules, triggers, and Chrome control via API/MCP need a paid plan. We didn't confirm paid plan prices.
❌ Not tested hands-on. Using it requires installing an extension and signing up for an account, which we didn't do for this issue.
How to wear it: Their own task guides include "Analyze a TikTok profile → post metrics + content brief" and "Research YouTube competitors." That's useful for gaming creators sizing up a niche without paying for a scraper.
The verdict: Free with no API key makes it the lowest-friction way to see what browser agents can do. Expect ads.
Not this season if: you hate sponsored content, need scheduled or automated runs on a free plan, or don't want an ad-supported agent acting inside your logged-in accounts.
THE MARGINS (not evaluated)
Google EmbeddingGemma 2 (open on-device multimodal embeddings, Apache 2.0) · CLI-Dispatch (hand Claude Code tasks off to Codex, Gemini, DeepSeek) · Offrun (manage multiple coding agents, Mac).
Reply with the tool you want us to test next.
— The Editors