21 harnesses, one hole

Description and use

<p>I read the intake evaluations for 21 agent harnesses. All of them solve the same problem: running an agent. None of them solve the problem of verifying one.</p> <p>That is the hole. Every harness on the market can launch a loop, fetch context, and emit code. None of them define "done" as something you can check. The survey that collected these 21 candidates is blunt about it: support must be mechanical, or it is theater.</p> <h2> The 21, in three tiers </h2> <p>The survey sorts the candidates by how Flow's discipline layer would actually integrate with them — not by marketing claims, but by

Problem

<p>I read the intake evaluations for 21 agent harnesses. All of them solve the same problem: running an agent. None of them solve the problem of verifying one.</p> <p>That is the hole. Every harness on the market can launch a loop, fetch context, and emit code. None of them define "done" as something you can check. The survey that collected these 21 candidates is blunt about it: support must be mechanical, or it is theater.</p> <h2> The 21, in three tiers </h2> <p>The survey sorts the candidates by how Flow's discipline layer would actually integrate with them — not by marketing claims, but by

Outcome

<p>I read the intake evaluations for 21 agent harnesses. All of them solve the same problem: running an agent. None of them solve the problem of verifying one.</p> <p>That is the hole. Every harness on the market can launch a loop, fetch context, and emit code. None of them define "done" as something you can check. The survey that collected these 21 candidates is blunt about it: support must be mechanical, or it is theater.</p> <h2> The 21, in three tiers </h2> <p>The survey sorts the candidates by how Flow's discipline layer would actually integrate with them — not by marketing claims, but by

Evidence

21 harnesses, one hole

Source