From 34dfbb66ef708acf9906a84b2ce4868b00cf110c Mon Sep 17 00:00:00 2001 From: pingqiu Date: Sun, 26 Apr 2026 20:57:51 -0700 Subject: [PATCH] =?UTF-8?q?v3-batch-process=20=C2=A715:=20per-agent=20acti?= =?UTF-8?q?on=20list=20at=20end=20of=20every=20response=20(cut=20user's=20?= =?UTF-8?q?routing=20load)?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit --- sw-block/design/v3-batch-process.md | 28 ++++++++++++++++++++++++++++ 1 file changed, 28 insertions(+) diff --git a/sw-block/design/v3-batch-process.md b/sw-block/design/v3-batch-process.md index 3ee889c91..2af2789cc 100644 --- a/sw-block/design/v3-batch-process.md +++ b/sw-block/design/v3-batch-process.md @@ -228,3 +228,31 @@ Based on what actually worked across T4 + G5: **Edge case — process changes (this doc):** QA proposes; architect signs (same as any scope-affecting change). **Edge case — hotfix-class** (per §6.3): sw can self-author + self-merge with QA spot-review. Architect ratifies if invariant-affecting. + +--- + +## §15 Inter-agent communication discipline + +Every response from QA, sw, or architect MUST end with an explicit per-agent action list. The user routes messages between agents — communication overhead is real cost. + +**Mandatory format at end of every substantive response:** + +``` +## Next actions + +| Agent | Action | +|---|---| +| **architect** | | +| **sw** | | +| **QA** | | +``` + +Rules: +1. Every agent gets a row, even if "none — standing by" +2. Actions must be **concrete + executable** ("commit X", "review Y", "decide path A vs B"), NOT vague ("consider Z") +3. If an action depends on another agent's prior action, name the dependency: "after sw lands G5-5, run -race ×10 on m01" +4. Skip the table only for genuinely no-decision-needed acknowledgements (e.g., one-line confirmations) + +**Why**: user's current load is ~90% copy-paste routing between agents. Explicit per-agent action lists at end of every response cut routing time. Every agent reads "what's MY action" without parsing a wall of text. + +This rule applies to **every agent's outputs**, not just QA's. Sw + architect responses to QA reviews must end with the same per-agent action list.