From b165bca96db91535189d31554cc1a87988c7f7e8 Mon Sep 17 00:00:00 2001 From: PAE Date: Thu, 12 Mar 2026 08:11:37 +0000 Subject: [PATCH] Adjudication: Task 3088135e-32b0-4ede-84fb-39e9506ebcde --- agents/iris/rag/agent.rag.md | 1 + 1 file changed, 1 insertion(+) diff --git a/agents/iris/rag/agent.rag.md b/agents/iris/rag/agent.rag.md index a84ca29..237cf68 100644 --- a/agents/iris/rag/agent.rag.md +++ b/agents/iris/rag/agent.rag.md @@ -12,3 +12,4 @@ - Policy: Initiate chapter editorial process by creating three specialized review tasks (developmental/structure, line edits, continuity) assigned to independent editors (Devon, Lane, Cora), followed by dependent roundtable debate for consensus. - Standard Practice: Simulate boardroom debates on narrative voice and power system consistency by spawning three parallel chapter review tasks (developmental/narrative by Devon, line/prose by Lane, continuity/worldbuilding by Cora) dependent on a synthesizing roundtable task. - Standard Practice: Chapter editorial reviews require three specialized agents—Devon (developmental), Lane (line), Cora (continuity)—with a dependent roundtable debate before consensus. +- Standard Practice: Simulate boardroom debates on narrative elements (e.g., voice consistency) by producing relevant content for review, then spawning three specialized editorial tasks—character/emotional analysis (Devon), world-building/magic mechanics (Lane), prose/pacing/tone (Cora)—followed by a dependent roundtable synthesis task.