Просмотр исходного кода

docs(agent): drop agent/stream-chunk from the two package READMEs

Codex convergence caught stale prose the doc-sync gate does not scan: the
dsh-agent README still listed `agent/stream-chunk` as an emit and the
dsh-agent-loop README still told UI plugins to listen on it. A plugin author
following either would silently get no chunks. Point both at the
`session/event` `assistant/chunk` feed (the agent-loop UI line also names the
`agent/*` control events a UI still uses).
Tianyi Cui 2 месяцев назад
Родитель
Сommit
e100c52154
2 измененных файлов с 1 добавлено и 2 удалено
  1. 1 1
      packages/core/agent-loop/README.md
  2. 0 1
      packages/core/agent/README.md

+ 1 - 1
packages/core/agent-loop/README.md

@@ -88,4 +88,4 @@ Everything that goes beyond "call the model, run the tools, repeat" belongs to p
 - Sandbox, permission, plan mode: `tools/pre-execute` (deny/ask gate), `tools/post-execute`
 - Sandbox, permission, plan mode: `tools/pre-execute` (deny/ask gate), `tools/post-execute`
 - Sub-agents: implemented outside the loop as `ctx.subagents` providers; in-process providers use `ctx.agents.create()` and owned `AgentHandle` teardown, while child streaming/progress and background/poll collection remain deferred.
 - Sub-agents: implemented outside the loop as `ctx.subagents` providers; in-process providers use `ctx.agents.create()` and owned `AgentHandle` teardown, while child streaming/progress and background/poll collection remain deferred.
 - Persistence: `session/event` + `session/flush`
 - Persistence: `session/event` + `session/flush`
-- UI: `agent/stream-chunk` + `agent/*` events
+- UI: `session/event` (assistant token stream, boundaries, tool activity) + `agent/*` control events (`agent/status`, `agent/created`/`agent/disposed`)

+ 0 - 1
packages/core/agent/README.md

@@ -52,7 +52,6 @@ Tool interception is the `tools/pre-execute` / `tools/post-execute` pair in [`ds
 
 
 #### Streaming + tool (emit)
 #### Streaming + tool (emit)
 
 
-- `agent/stream-chunk` — raw chunk from the model (token-level UI/log feed)
 - `agent/steering` — steering content injected mid-turn
 - `agent/steering` — steering content injected mid-turn
 - `agent/error` — step/turn error
 - `agent/error` — step/turn error