pi-openai-toolkit 0.20.7 → 0.20.8

package/docs/configuration.md

@@ -63,7 +63,7 @@ Place `context` under `defaults` or an exact `models` entry.
 | `remoteCompaction.allowContinuityBreak` | `false` | Allow restarting from Pi context after a foreign compaction entry. It does not make malformed opaque checkpoints replayable. |
 | `remoteCompaction.apis` | `["openai-responses", "openai-codex-responses", "azure-openai-responses"]` | May narrow this set; `[]` permits no remote-compaction API. Azure uses Pi's resolved `AZURE_OPENAI_*` environment values for the `/openai/v1` Responses path, deployment mapping, and `api-version`; unsupported entries are errors. |
 | `nativeFallback.enabled` | `true` | Enable the existing native-method fallback tier. |
-| `nativeFallback.model` | `null` | Optional summary-model reference on the remote-ineligible path. After a remote attempt fails, the existing producer-first fallback order is retained. |
+| `nativeFallback.model` | `null` | Summary-model reference on the remote-ineligible path. Character-count estimates do not bypass it; resolution, auth or summary failure cancels compaction without retrying the active model. After a remote attempt fails, the existing producer-first selection is retained. Disabled/unset/same-model selections keep Pi's default path. |
 | `nativeFallback.thinkingLevel` | `"off"` | `off`, `minimal`, `low`, `medium`, `high`, `xhigh`, or `max`; actual model support still applies. |
 | `remoteWindows.leaveManagedMode` | `"warn"` | Legacy compatibility field. Both values preserve the same behavior: non-window models receive only the current local window, and no bulk close-out compaction is started for retired Remote Context history. |
 | `remoteWindows.reminderThresholdPercent` | `5` | Integer 0-100. `0` disables both the once-per-window reminder and exhausted-window fallback. |

package/docs/internals.md

@@ -97,7 +97,7 @@ State invariants the lifecycle code must keep honest:
 - Window identity and cleanup eligibility are separate. A persisted `codex-context-window` message restores identity after reload/fork, but its historical `trimPreviousWindow` flag is not an outstanding command. A newer native summary or opaque checkpoint supersedes old cleanup without erasing window identity. Missing or edited boundary evidence is never repaired by loading raw history.
 - Maintenance compaction and local-consumer detachment use the same effective-boundary rules. Only a safe current boundary can produce a no-summary compaction; obsolete cleanup must not move the retained start backwards or replace a newer summary/checkpoint. Preparation alone changes no durable state, and cancellation, retry and branch navigation are resolved from current evidence rather than callback acknowledgments. User manual `/compact` still follows the managed handoff below.
 - A non-window consumer receives only the effective current context, with the system/tool head preserved and internal window markers removed. When detachment is needed, no compaction producer receives retired Remote history. Once a later ordinary compaction owns effective context, the historical window does not suppress its summary or valid opaque replay. A genuinely new window can retire that checkpoint. Disabling opaque checkpoint generation through `responsesApis` does not disable required replay of an inherited checkpoint; an unsupported transport, authentication failure, or endpoint mismatch aborts the request instead of exposing its placeholder. The legacy `leaveManagedMode` setting remains compatible but does not start a bulk close-out from `model_select` or `agent_settled`.
-- The configured native-fallback summary model is still checked against the physical request size (`estimateSummarizationRequest`). The guard remains a provider-safety invariant; the Remote-window fix avoids the oversized request by excluding managed history before any compaction producer can be called, rather than removing the guard or chunking the request.
+- Native text summarization delegates prompt construction, tool-result truncation, split-turn requests and output limits to Pi's `compact()`. Toolkit does not use character-count estimates to reject the configured summary model. If its resolution, authentication or summary generation fails, compaction is cancelled with the error and context is preserved; the active model is not an implicit retry target. Disabled/unset/same-model selections retain Pi's default path. Remote-window projection still excludes retired managed history before any producer call.
 
 ---
 

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "pi-openai-toolkit",
-  "version": "0.20.7",
+  "version": "0.20.8",
   "private": false,
   "type": "module",
   "engines": {

package/src/extension-runtime.ts

@@ -926,8 +926,6 @@ async function handleSessionBeforeCompact(
 			reason: fallback.reason,
 			modelSpec: fallback.modelSpec,
 			errorMessage: fallback.errorMessage,
-			...(fallback.estimatedTokens === undefined ? {} : { estimatedTokens: fallback.estimatedTokens }),
-			...(fallback.contextWindow === undefined ? {} : { contextWindow: fallback.contextWindow }),
 		},
 		config,
 		ctx,
@@ -938,19 +936,14 @@ async function handleSessionBeforeCompact(
 		fallback.reason === "disabled" ||
 		fallback.reason === "no-model-configured" ||
 		fallback.reason === "same-as-current-model";
-	if (fallback.reason === "model-window-too-small") {
-		// The configured summary model is narrower than the request Pi would send for it.
-		// Skipping is the success path here: pi's current model gets one chance instead of
-		// the provider terminating the oversized summary.
+	if (!intentionalSkip) {
+		// An explicit summary model is authoritative: the active model may be
+		// unable to summarize at all. Preserve context on failure, never reroute.
 		notifyWarning(
 			ctx,
-			`compaction summary model "${fallback.modelSpec}" cannot fit ~${fallback.estimatedTokens} tokens in its ${fallback.contextWindow} window; compacting with the current model instead`,
-		);
-	} else if (!intentionalSkip) {
-		notifyWarning(
-			ctx,
-			`compaction model "${fallback.modelSpec}" unusable (${fallback.reason}${fallback.errorMessage ? `: ${fallback.errorMessage}` : ""}); using pi's default compaction`,
+			`compaction model "${fallback.modelSpec ?? fallbackModelSpec}" failed (${fallback.reason}${fallback.errorMessage ? `: ${fallback.errorMessage}` : ""}); compaction cancelled, context preserved; not switching to the current model`,
 		);
+		return { cancel: true };
 	}
 
 	// Branch 3: pi's default native compaction with the current model.

package/src/native-fallback.ts

@@ -1,8 +1,5 @@
 import {
 	compact,
-	convertToLlm,
-	DEFAULT_COMPACTION_SETTINGS,
-	serializeConversation,
 	type CompactionResult,
 	type ExtensionContext,
 	type SessionBeforeCompactEvent,
@@ -23,9 +20,7 @@ export type NativeFallbackFailureReason =
 	| "auth-failed"
 	| "aborted"
 	| "empty-summary"
-	| "compact-failed"
-	/** The configured summary model cannot fit the request it would have to send. */
-	| "model-window-too-small";
+	| "compact-failed";
 
 export type NativeFallbackResult =
 	| {
@@ -38,8 +33,6 @@ export type NativeFallbackResult =
 			reason: NativeFallbackFailureReason;
 			modelSpec?: string;
 			errorMessage?: string;
-			estimatedTokens?: number;
-			contextWindow?: number;
 	  };
 
 /** pi's exported native compact(); injectable for tests. */
@@ -60,116 +53,6 @@ function toErrorMessage(error: unknown): string {
 	return error instanceof Error ? error.message : String(error);
 }
 
-/** Summarization instructions plus the framing around the serialized transcript. */
-const SUMMARIZATION_TEMPLATE_TOKENS = 1024;
-const HISTORY_OUTPUT_RESERVE_RATIO = 0.8;
-const TURN_PREFIX_OUTPUT_RESERVE_RATIO = 0.5;
-
-export type SummarizationRequestSize = {
-	inputTokens: number;
-	outputTokens: number;
-	totalTokens: number;
-};
-
-function estimateTextTokens(text: string | undefined): number {
-	return text ? Math.ceil(String(text).length / 4) : 0;
-}
-
-function estimateSerializedConversationTokens(
-	messages: SessionBeforeCompactEvent["preparation"]["messagesToSummarize"],
-): number {
-	const serialized = serializeConversation(convertToLlm(messages));
-	return estimateTextTokens(serialized);
-}
-
-function getEffectiveReserveTokens(preparation: SessionBeforeCompactEvent["preparation"]): number {
-	const reserveTokens = preparation.settings?.reserveTokens;
-	return typeof reserveTokens === "number" && Number.isFinite(reserveTokens) && reserveTokens >= 0
-		? reserveTokens
-		: DEFAULT_COMPACTION_SETTINGS.reserveTokens;
-}
-
-function getOutputReserveTokens(
-	reserveTokens: number,
-	model: { maxTokens?: number },
-	ratio: number,
-): number {
-	const outputTokens = Math.floor(ratio * reserveTokens);
-	return typeof model.maxTokens === "number" && model.maxTokens > 0
-		? Math.min(outputTokens, model.maxTokens)
-		: outputTokens;
-}
-
-function estimateSummaryRequest(
-	messages: SessionBeforeCompactEvent["preparation"]["messagesToSummarize"],
-	outputTokens: number,
-	previousSummary?: string,
-	customInstructions?: string,
-): SummarizationRequestSize {
-	const inputTokens =
-		SUMMARIZATION_TEMPLATE_TOKENS
-		+ estimateSerializedConversationTokens(messages)
-		+ estimateTextTokens(previousSummary)
-		+ estimateTextTokens(customInstructions);
-	return { inputTokens, outputTokens, totalTokens: inputTokens + outputTokens };
-}
-
-/**
- * Size of the individual request(s) Pi's `compact()` is about to send.
- *
- * Pi first converts messages with `convertToLlm()` and then applies
- * `serializeConversation()`, which truncates tool results before they enter the
- * summarization prompt. Split-turn compaction sends history and turn-prefix
- * summaries as separate requests, so this returns the largest individual request
- * rather than adding their inputs together. The fixed allowance remains
- * deliberately conservative for Pi's system prompt, tags, and instructions.
- *
- * A session driven by a wide-window model (400k) can produce a preparation that a
- * narrower summary model (272k) cannot accept. Callers must not attempt that request;
- * a missing `contextWindow` skips the check instead of guessing.
- */
-export function estimateSummarizationRequest(
-	preparation: SessionBeforeCompactEvent["preparation"],
-	model: { contextWindow?: number; maxTokens?: number },
-	customInstructions?: string,
-): SummarizationRequestSize {
-	const reserveTokens = getEffectiveReserveTokens(preparation);
-	const historyMessages = preparation.messagesToSummarize ?? [];
-	const turnPrefixMessages = preparation.turnPrefixMessages ?? [];
-	const historyOutputTokens = getOutputReserveTokens(reserveTokens, model, HISTORY_OUTPUT_RESERVE_RATIO);
-	const historyRequest = estimateSummaryRequest(
-		historyMessages,
-		historyOutputTokens,
-		preparation.previousSummary,
-		customInstructions,
-	);
-
-	if (!preparation.isSplitTurn || turnPrefixMessages.length === 0) {
-		return historyRequest;
-	}
-
-	const requests = [
-		...(historyMessages.length > 0 ? [historyRequest] : []),
-		estimateSummaryRequest(
-			turnPrefixMessages,
-			getOutputReserveTokens(reserveTokens, model, TURN_PREFIX_OUTPUT_RESERVE_RATIO),
-		),
-	];
-	return requests.reduce((largest, request) =>
-		request.totalTokens > largest.totalTokens ? request : largest,
-	);
-}
-
-/** Whether the summary model can accept this request at all; unknown windows always pass. */
-export function fitsSummarizationRequest(
-	size: SummarizationRequestSize,
-	model: { contextWindow?: number },
-): boolean {
-	const contextWindow = model.contextWindow;
-	if (typeof contextWindow !== "number" || contextWindow <= 0) return true;
-	return size.totalTokens <= contextWindow;
-}
-
 /**
  * Run pi's native compaction method with a caller-resolved summary model.
  *
@@ -218,20 +101,9 @@ export async function runNativeFallbackCompaction(args: {
 		return { ok: false, reason: "same-as-current-model", modelSpec: spec };
 	}
 
-	const size = estimateSummarizationRequest(
-		event.preparation,
-		model as { contextWindow?: number; maxTokens?: number },
-		event.customInstructions,
-	);
-	if (!fitsSummarizationRequest(size, model as { contextWindow?: number })) {
-		return {
-			ok: false,
-			reason: "model-window-too-small",
-			modelSpec: spec,
-			estimatedTokens: size.totalTokens,
-			contextWindow: (model as { contextWindow?: number }).contextWindow,
-		};
-	}
+	// Pi owns prompt construction, tool-result truncation and split-turn requests.
+	// A chars/4 estimate is not target-model token accounting and must not veto
+	// the user's summary model. Surface actual provider failures to the caller.
 
 	let auth: ResolvedAuth;
 	try {