From f13fb2e940c704f68b3fcbe782de299ec98dad48 Mon Sep 17 00:00:00 2001 From: Dmitry Golubev Date: Sat, 9 May 2026 12:04:56 +0300 Subject: [PATCH] docs: update stale sub-agent context injection details (#79470) Summary: - Correct stale sub-agent docs to list the current minimal bootstrap context files. - Add changelog credit for @lastguru-net. Verification: - pnpm docs:list - pnpm exec oxfmt --check --threads=1 docs/tools/subagents.md CHANGELOG.md - git diff --check - GitHub CI: check-docs, Real behavior proof --- CHANGELOG.md | 1 + docs/tools/subagents.md | 2 +- 2 files changed, 2 insertions(+), 1 deletion(-) diff --git a/CHANGELOG.md b/CHANGELOG.md index 36937b32ad0..244e90be316 100644 --- a/CHANGELOG.md +++ b/CHANGELOG.md @@ -145,6 +145,7 @@ Docs: https://docs.openclaw.ai ### Fixes +- Docs/Subagents: correct the listed sub-agent bootstrap context files to include `SOUL.md`, `IDENTITY.md`, and `USER.md`. (#79470) Thanks @lastguru-net. - OpenAI/Codex: install the Codex runtime plugin from npm during OpenAI onboarding and load it automatically for implicit OpenAI model routes, while preserving manual PI runtime overrides. Fixes #79358. - OpenAI/realtime voice: defer `response.create` while a realtime response is still active, retry after `response.done`/`response.cancelled`, and align GA input transcription/noise-reduction defaults with the Codex realtime reference so Discord/Voice Call consult results can resume speaking instead of tripping the active-response race. - OpenAI/realtime voice: avoid duplicate barge-in cancellation requests, log realtime model interruption/cutoff events in Discord voice logs, and treat OpenAI's no-active-response cancellation reply as a completed cancel so Discord voice sessions do not wedge pending speech after fast interruptions. diff --git a/docs/tools/subagents.md b/docs/tools/subagents.md index 99584b9c13b..f7f484922ae 100644 --- a/docs/tools/subagents.md +++ b/docs/tools/subagents.md @@ -547,7 +547,7 @@ still need normal device approval for scope upgrades. - Sub-agent announce is **best-effort**. If the gateway restarts, pending "announce back" work is lost. - Sub-agents still share the same gateway process resources; treat `maxConcurrent` as a safety valve. - `sessions_spawn` is always non-blocking: it returns `{ status: "accepted", runId, childSessionKey }` immediately. -- Sub-agent context only injects `AGENTS.md` + `TOOLS.md` (no `SOUL.md`, `IDENTITY.md`, `USER.md`, `HEARTBEAT.md`, or `BOOTSTRAP.md`). +- Sub-agent context only injects `AGENTS.md`, `TOOLS.md`, `SOUL.md`, `IDENTITY.md` and `USER.md` (no `MEMORY.md`, `HEARTBEAT.md`, or `BOOTSTRAP.md`). - Maximum nesting depth is 5 (`maxSpawnDepth` range: 1–5). Depth 2 is recommended for most use cases. - `maxChildrenPerAgent` caps active children per session (default `5`, range `1–20`).