From d77c1424889105c459ff268c0d94e2ca38352816 Mon Sep 17 00:00:00 2001 From: Don Garrett Date: Fri, 2 Oct 2026 18:35:00 +0000 Subject: [PATCH] Accept a plain-language go-ahead, announced and revision-bound Co-Authored-By: Claude Opus 5.5 (1M context) Claude-Session: https://claude.ai/code/session_01PPQ14L3LikyefCMFBqEbKE --- docs/Architecture.md | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-) diff --git a/docs/Architecture.md b/docs/Architecture.md index 45264c4..7dc67b1 100644 --- a/docs/Architecture.md +++ b/docs/Architecture.md @@ -396,6 +396,12 @@ that works in every state: `status`, `stop`, `abandon`, `approve [rev-N]`, `go` drawn from the work item's context, and is never read as approval, rejection or a command. How a question is recognised (explicit marker or classification) is an open question (Q); a misread must default to asking, never to acting. +- **One exception: a plain-language go-ahead.** A conversational reply to a + restatement ("yep, that's it") may be taken as `go` (D1), because a wrong reading + costs only time: the work is contained and ends in a PR. The agent says so before + acting ("taking that as a go-ahead on rev-2; starting"), and the reading is bound + to the revision up when the reply arrived (L3). Approvals of anything else (ops + step sequences, class-3 actions) still need an explicit `approve`. - **Everything the operator says is recorded** in the work item's transcript (L7), even when no action follows. @@ -811,7 +817,8 @@ request to mean, how it fits the architecture, and its first step. Then: on the operator, so the agent may work on something else; when it expires the item is ready. - **Anything else** (a feature, a change, a question) waits for the operator's - go-ahead after the restatement; the conversation is where the design is agreed. + go-ahead after the restatement: `go`, or a plain-language yes, which the agent + announces it is acting on (L5). The conversation is where the design is agreed. - **Talk first when:** the issue is labelled `discuss`; the agent finds the report unclear (it asks its questions instead of restating); or it can't reproduce a bug. It never fixes a bug it couldn't see.