Add an authored headless scenario that appends a changed prompt and then replaces a surface node before the third step. The unchanged-rendering series start must keep both prompt versions and write no system/message event; the scenario fails against the pre-fix projection. Also tighten the tool-series-start assertions and make the prefix premise explicit.
An in-history route carries prompt updates as later system nodes while node 0 stays stale. A surface replacement (compaction, tool-result prune, image offload) starts a new request series, and the projection used to re-baseline node 0 even when the rendering already matched the latest surviving node: it emptied that node and wrote identical text into node 0, so the request differed at its first message and the provider prefix cache missed from the first token (#6973).
Re-baseline node 0 only when the effective text must move: an incapable route, an empty rendering, or a series start whose latest surviving non-empty node differs from the rendering. A capable series start with matching text emits no system event; the loop still logs the request series and rebuilds declarations. Retry compaction still folds the prompt to the head when it shadows the latest node.
A compaction replacement after an appended in-history prompt starts a new request series, and the projection folds the prompt back into node 0 although the effective text is unchanged. The request moves the system segment to wire message 0, so the provider prefix cache misses from the first token (#6973).
The regression fails before the fix: the third request carries one system text ("prompt two") instead of keeping the appended node behind the cached prefix.
ctx.tools.register(definition, { when }) adds a synchronous or asynchronous
condition. AgentLoop resolves conditions once per proposed step with the
agent, the model selection that step preparation captured, and the turn
signal, before prompt assembly and agent/pre-step. The resulting
ResolvedToolSet governs tool declarations, Native dispatch, PTC bindings,
and nested calls for that step and its retries, and the turn disposes it
when the proposal is rejected or fails, when no step opens, and after the
step. A false condition removes the tool everywhere; a thrown condition
ends the turn with TOOL_ACTIVATION_FAILED. Calls whose registration was
replaced or removed during the step return UNKNOWN_TOOL.
system-prompt/assemble now transforms PromptPresentation without a tool
array, so tool declarations stay provider-owned. exit_plan_mode is declared
only for steps whose selected plan state is active, the same state that
renders plan guidance; it still rejects a call before the selection is
logged.
Model selection stays as master defines it; this change adds no model
selection event or Agent field.
Add @deepseek-ai/dsh-agent-model. It provides ctx.agentModel
(resolveForNextStep, select, applyConfiguredRoute) and owns the
modelSelection projection, which the Agent registry no longer registers.
dsh-agent-loop now injects agentModel; the base and SDK-minimal bundles
and the Python SDK runtime include it.
The provider keeps the projection registered until consumers that are
still unloading finish their cleanup. The model-switch notice becomes an
opt-in per-Agent plugin at @deepseek-ai/dsh-agent-model/notice, and
agent/request now also rejects reasoning-effort changes after step
preparation.
Mount the provider in every composition that loads dsh-agent-loop, and
update consumers, the switch-notice snapshot driver, docs, and the
upgrade guide.
Master removed `backgroundMode` and the `run_in_background` argument, so
every delegation starts a managed activation. Mount the working-directory
owner SubagentRuntime injects and the JSONL persistence root the spawn
provider requires, and synchronize on the child's cancellation, which the
parent teardown or the manager drain raises instead of a parent-only cause.
