A separate evaluator agent is what makes long agent runs worth the money

Mon, 05 Oct 2026 17:00:00 +0000

Ask one agent to build a whole application and two things tend to go wrong. It loses the thread as its context fills, and it grades its own work generously. Anthropic’s engineering team wrote up a harness design for long-running application development that attacks both, and the numbers are honest about what it costs.

What 16 parallel agents building a C compiler teach about running agents at scale

Sun, 04 Oct 2026 12:00:00 +0000

Anthropic’s engineering team handed 16 Claude instances a goal, a shared repo and a test suite, and let them write a C compiler in Rust. The result, described in an engineering post , is a roughly 100,000-line compiler that can build Linux 6.9 on x86, ARM and RISC-V, plus projects like SQLite, postgres and redis. It took nearly two weeks, close to 2,000 Claude Code sessions, 2 billion input tokens and 140 million output tokens, for just under $20,000.

Harness on Tokenise

A separate evaluator agent is what makes long agent runs worth the money

What 16 parallel agents building a C compiler teach about running agents at scale