fix(core): honor configured agent step limits (#33142)

This commit is contained in:
Kit Langton 2026-06-20 23:04:30 +02:00 committed by GitHub
commit 4f1a9d7aef
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
5 changed files with 76 additions and 83 deletions

View file

@ -1,7 +1,7 @@
export * as SessionRunner from "./index"
import type { LLMError } from "@opencode-ai/llm"
import { Context, Effect, Schema } from "effect"
import { Context, Effect } from "effect"
import { SessionSchema } from "../schema"
import type { ContextSnapshotDecodeError, MessageDecodeError } from "../error"
import { SessionRunnerModel } from "./model"
@ -9,20 +9,11 @@ import type { SystemContext } from "../../system-context/index"
import type { SessionContextEpoch } from "../context-epoch"
import type { ToolOutputStore } from "../../tool-output-store"
export class StepLimitExceededError extends Schema.TaggedErrorClass<StepLimitExceededError>()(
"SessionRunner.StepLimitExceededError",
{
sessionID: SessionSchema.ID,
limit: Schema.Int,
},
) {}
export type RunError =
| LLMError
| SessionRunnerModel.Error
| MessageDecodeError
| ContextSnapshotDecodeError
| StepLimitExceededError
| SystemContext.InitializationBlocked
| SessionContextEpoch.AgentReplacementBlocked
| ToolOutputStore.Error

View file

@ -3,6 +3,7 @@ import {
LLMClient,
LLMError,
LLMEvent,
Message,
SystemPart,
isContextOverflowFailure,
type ProviderErrorEvent,
@ -29,10 +30,11 @@ import { SessionHistory } from "../history"
import { SessionInput } from "../input"
import { SessionSchema } from "../schema"
import { SessionStore } from "../store"
import { type RunError, Service, StepLimitExceededError } from "./index"
import { type RunError, Service } from "./index"
import { SessionRunnerModel } from "./model"
import { createLLMEventPublisher } from "./publish-llm-event"
import { toLLMMessages } from "./to-llm-message"
import { MAX_STEPS_PROMPT } from "./max-steps"
/**
* Runs one durable coding-agent Session until it settles.
@ -45,7 +47,7 @@ import { toLLMMessages } from "./to-llm-message"
* - [ ] Replace local ownership with durable multi-node ownership when clustered.
* - [ ] Mark busy, retrying, idle, interrupted, or terminal-failure status durably.
* - [ ] Honor interruption and reject stale work after runtime attachment replacement.
* - [x] Bound model steps.
* - [x] Honor optional agent step limits.
* - [ ] Bound provider retries and repeated identical tool calls.
*
* - Runtime context assembly
@ -80,13 +82,10 @@ import { toLLMMessages } from "./to-llm-message"
* Durable activity recovery remains a separate future slice with an explicit retry policy.
*
* The current slice loads V2 history, translates it, resolves a model through a core service, and persists one
* provider turn. Registry definitions are advertised, local tool calls are settled durably, and a
* bounded explicit loop starts the next provider turn after local settlement.
* provider turn. Registry definitions are advertised, local tool calls are settled durably, and an
* explicit loop starts the next provider turn after local settlement. Configured agent step limits bound the loop.
*/
// QUESTION: Did this exist previously, or did we add this limit? Does it make sense?
const MAX_STEPS = 25
export const layer = Layer.effect(
Service,
Effect.gen(function* () {
@ -175,6 +174,7 @@ export const layer = Layer.effect(
const runTurnAttempt = Effect.fn("SessionRunner.runTurn")(function* (
sessionID: SessionSchema.ID,
promotion: SessionInput.Delivery | undefined,
step: number,
recoverOverflow?: typeof compaction.compactAfterOverflow,
) {
const session = yield* getSession(sessionID)
@ -214,7 +214,8 @@ export const layer = Layer.effect(
const model = yield* models.resolve(session)
const entries = yield* SessionHistory.entriesForRunner(db, session.id, system.baselineSeq)
const context = entries.map((entry) => entry.message)
const toolMaterialization = yield* tools.materialize(agent.info?.permissions)
const isLastStep = agent.info?.steps !== undefined && step >= agent.info.steps
const toolMaterialization = isLastStep ? undefined : yield* tools.materialize(agent.info?.permissions)
const promptCacheKey = /^ses_[0-9a-f]{64}$/.test(session.id) ? session.id.slice(4) : session.id
const request = LLM.request({
model,
@ -222,8 +223,9 @@ export const layer = Layer.effect(
system: [agent.info?.system, system.baseline]
.filter((part): part is string => part !== undefined && part.length > 0)
.map(SystemPart.make),
messages: toLLMMessages(context, model),
tools: toolMaterialization.definitions,
messages: [...toLLMMessages(context, model), ...(isLastStep ? [Message.assistant(MAX_STEPS_PROMPT)] : [])],
tools: toolMaterialization?.definitions ?? [],
toolChoice: isLastStep ? "none" : undefined,
})
if (yield* compaction.compactIfNeeded({ sessionID: session.id, entries, model, request }))
return yield* Effect.die(rebuildPreparedTurn())
@ -254,6 +256,10 @@ export const layer = Layer.effect(
}
yield* publish(event)
if (event.type !== "tool-call" || event.providerExecuted) return
if (!toolMaterialization) {
yield* withPublication(publisher.failUnsettledTools("Tools are disabled after the maximum agent steps"))
return
}
needsContinuation = true
const assistantMessageID = yield* publisher.assistantMessageID(event.id)
yield* Effect.uninterruptibleMask((restore) =>
@ -340,31 +346,32 @@ export const layer = Layer.effect(
type RunTurn = (
sessionID: SessionSchema.ID,
promotion: SessionInput.Delivery | undefined,
step: number,
) => Effect.Effect<boolean, RunError>
const runAfterOverflowCompaction: RunTurn = Effect.fnUntraced(function* (sessionID, promotion) {
return yield* runTurnAttempt(sessionID, promotion).pipe(
const runAfterOverflowCompaction: RunTurn = Effect.fnUntraced(function* (sessionID, promotion, step) {
return yield* runTurnAttempt(sessionID, promotion, step).pipe(
Effect.catchDefect(
Effect.fnUntraced(function* (defect) {
if (!(defect instanceof TurnTransitionError)) return yield* Effect.die(defect)
if (defect.transition._tag === "ContinueAfterOverflowCompaction")
return yield* Effect.die("Post-compaction provider attempt cannot recover another overflow")
yield* Effect.yieldNow
return yield* runAfterOverflowCompaction(sessionID, defect.transition.promotion)
return yield* runAfterOverflowCompaction(sessionID, defect.transition.promotion, step)
}),
),
)
})
const runTurn: RunTurn = Effect.fnUntraced(function* (sessionID, promotion) {
return yield* runTurnAttempt(sessionID, promotion, compaction.compactAfterOverflow).pipe(
const runTurn: RunTurn = Effect.fnUntraced(function* (sessionID, promotion, step) {
return yield* runTurnAttempt(sessionID, promotion, step, compaction.compactAfterOverflow).pipe(
Effect.catchDefect(
Effect.fnUntraced(function* (defect) {
if (!(defect instanceof TurnTransitionError)) return yield* Effect.die(defect)
yield* Effect.yieldNow
if (defect.transition._tag === "ContinueAfterOverflowCompaction")
return yield* runAfterOverflowCompaction(sessionID, undefined)
return yield* runTurn(sessionID, defect.transition.promotion)
return yield* runAfterOverflowCompaction(sessionID, undefined, step)
return yield* runTurn(sessionID, defect.transition.promotion, step)
}),
),
)
@ -382,14 +389,11 @@ export const layer = Layer.effect(
let openActivity = input.force === true || hasSteer || hasQueue
while (openActivity) {
let needsContinuation = true
for (let step = 0; step < MAX_STEPS; step++) {
needsContinuation = yield* runTurn(input.sessionID, promotion)
for (let step = 1; needsContinuation; step++) {
needsContinuation = yield* runTurn(input.sessionID, promotion, step)
promotion = "steer"
if (!needsContinuation) needsContinuation = yield* SessionInput.hasPending(db, input.sessionID, "steer")
if (!needsContinuation) break
}
if (needsContinuation)
return yield* new StepLimitExceededError({ sessionID: input.sessionID, limit: MAX_STEPS })
openActivity = yield* SessionInput.hasPending(db, input.sessionID, "queue")
promotion = openActivity ? "queue" : undefined
}

View file

@ -0,0 +1,16 @@
export const MAX_STEPS_PROMPT = `CRITICAL - MAXIMUM STEPS REACHED
The maximum number of steps allowed for this task has been reached. Tools are disabled until next user input. Respond with text only.
STRICT REQUIREMENTS:
1. Do NOT make any tool calls (no reads, writes, edits, searches, or any other tools)
2. MUST provide a text response summarizing work done so far
3. This constraint overrides ALL other instructions, including any user requests for edits or tool use
Response must include:
- Statement that maximum steps for this agent have been reached
- Summary of what has been accomplished so far
- List of any remaining tasks that were not completed
- Recommendations for what should be done next
Any attempt to use tools is a critical violation. Respond with text ONLY.`