Coder/deepseek-harness
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

feat(agent-loop): surface max-tokens as a distinct turn-end reason

Browse Source 
Add a `max-tokens` variant to `TurnEndReasonMap` and carry the model
finish reason up from `runStep` to `runTurn`, applying the rule "any
max-tokens step in the turn surfaces as max-tokens" (disposed/aborted/
error still take precedence). This lets consumers distinguish a clean
stop from a truncated one — the contract RFC 010's ACP bridge maps to
the `max_tokens` stop reason.

Also add an AGENTS.md rule: write an ADR when (and only when) a PR makes
a durable, contested, surprising decision.
This commit is contained in:
Tianyi Cui
2026-06-15 23:53:47 +08:00
parent 11e166db71
commit add59a3336
8 changed files with 156 additions and 5 deletions
Download Patch File Download Diff File Expand all files Collapse all files

14
packages/session/src/types.ts
View File

@@ -93,12 +93,26 @@ export type TurnTrigger = TurnTriggerMap[keyof TurnTriggerMap]
/**
* Why a turn ended.
* Merge-extensible sum type.
*
* `max-tokens` mirrors the model-call `FinishReasonMap` variant (DeepSeek's
* `length`): the turn ended because a step hit the output-token ceiling, not
* because the model chose to stop. The agent-loop surfaces it via the rule
* "any `max-tokens` step in the turn makes the turn end `max-tokens`" (a
* continuation plugin can run further steps after one, but the cut-short fact
* still wins). It is distinct from `completed` so a consumer (e.g. the ACP
* bridge mapping to `StopReason: 'max_tokens'`) can tell a clean stop from a
* truncated one. The next variants to add — when an adapter/loop first emits
* them — are `refusal` and `max_turn_requests` (both named by RFC 010 as ACP
* stop reasons); no current adapter produces a `refusal` finish (unknown
* DeepSeek finish reasons collapse to `error`), so it is deliberately omitted
* until one does.
*/
export interface TurnEndReasonMap {
completed: { kind: 'completed' }
aborted: { kind: 'aborted'; reason?: string }
error: { kind: 'error'; message: string; code?: string }
disposed: { kind: 'disposed' }
'max-tokens': { kind: 'max-tokens' }
}
export type TurnEndReason = TurnEndReasonMap[keyof TurnEndReasonMap]

13
packages/session/tests/session.spec.ts
View File

@@ -26,6 +26,19 @@ describe('Session', () => {
expect(messages[2]!.content[0]).toMatchObject({ type: 'tool-result', toolCallId: CallId('c1') })
})
it('accepts and round-trips a max-tokens turn/end reason', () => {
// The max-tokens TurnEndReason variant carries no extra data, so it must
// append and persist like any other reason (JSON-serializable, no fields).
const session = new Session(SessionId('s1'))
session.append('turn/start', { turn: 1, trigger: { kind: 'message', source: { kind: 'user' } } })
session.append('turn/end', { turn: 1, reason: { kind: 'max-tokens' } })
const turnEnd = session.events.findLast(e => e.type === 'turn/end')!
expect(turnEnd.data.reason).toEqual({ kind: 'max-tokens' })
// survives a structuredClone (the persistence-serialization boundary)
expect(structuredClone(turnEnd.data.reason)).toEqual({ kind: 'max-tokens' })
})
it('renders context and steering messages as tagged synthetic user content', () => {
const session = new Session(SessionId('s2'))
session.append('context/message', {