The post refers to a Claude Code mod where Jev selects the agent and model for each subagent call, with a three-round benchmark write-up on the author's blog. The tweet serves as a discussion thread and only links to the blog post, without repository, install steps, or demo media.
I wrote up the full benchmark for my Claude Code mod that lets Jev pick the agent and model for every subagent call. Three rounds, and the first two lost. If you're routing subagents to cheaper models yourself, or you think my 0.8 threshold is wrong, reply here. This is the thread the post links to for discussion. https://benenewton.com/blog/jev-agent-routing-claude-code?utm_source=twitter&utm_medium=social&utm_campaign=jev-routing-post-discussion-tweet
The tweet by Ben Newton links to his blog post containing a full benchmark for a Claude Code mod that lets Jev pick the agent and model for every subagent call, noting three rounds with the first two losing. It invites replies from users routing subagents to cheaper models or disagreeing with the 0.8 threshold, and acts as the discussion thread for the linked article. No repository link, installation instructions, or demo media are included in the post itself.
See the original source for installation instructions.
I wrote up the full benchmark for my Claude Code mod that lets Jev pick the agent and model for every subagent call. Three rounds, and the first two lost.
If you're routing subagents to cheaper models yourself, or you think my 0.8 threshold is wrong, reply here. This is the thread the post links to for discussion.
https://benenewton.com/blog/jev-agent-routing-claude-code?utm_source=twitter&utm_medium=social&utm_campaign=jev-routing-post-discussion-tweet
Session-as-skill browser automation plugin for Claude Code: Playwright-backed MCP tools, RVF cognitive containers, AgentDB selector/cookie memory, AIDefence gates, plus a tighten-only tool guard mod.
Gives Claude a register_assumption tool and shows the turn's assumptions, decisions and considered-but-not-done items above the prompt
Test wrapper mod for hermitd, an always-on Claude Code agent plugin with chat channels, routines, heartbeats, and a learning loop.