@oh-my-pi/pi-coding-agent 16.1.10 → 16.1.12

package/src/hindsight/backend.ts

@@ -15,7 +15,7 @@ import type { AgentSession } from "../session/agent-session";
 import { type BankScope, computeBankScope } from "./bank";
 import { createHindsightClient } from "./client";
 import { isHindsightConfigured, loadHindsightConfig } from "./config";
-import type { HindsightMessage } from "./content";
+import { type HindsightMessage, hasSubstantiveContent } from "./content";
 import { HindsightSessionState } from "./state";
 
 const STATIC_INSTRUCTIONS = [
@@ -330,7 +330,7 @@ function flattenMessagesForRecall(messages: AgentMessage[]): HindsightMessage[]
 		if (msg.role === "user") {
 			const content = msg.content;
 			if (typeof content === "string") {
-				if (content.trim()) out.push({ role: "user", content });
+				if (hasSubstantiveContent(content)) out.push({ role: "user", content });
 				continue;
 			}
 			if (Array.isArray(content)) {
@@ -338,7 +338,7 @@ function flattenMessagesForRecall(messages: AgentMessage[]): HindsightMessage[]
 					.filter((b): b is { type: "text"; text: string } => !!b && (b as { type?: unknown }).type === "text")
 					.map(b => b.text)
 					.join("\n");
-				if (text.trim()) out.push({ role: "user", content: text });
+				if (hasSubstantiveContent(text)) out.push({ role: "user", content: text });
 			}
 			continue;
 		}
@@ -347,7 +347,7 @@ function flattenMessagesForRecall(messages: AgentMessage[]): HindsightMessage[]
 				.filter((b): b is { type: "text"; text: string } => b.type === "text")
 				.map(b => b.text)
 				.join("\n");
-			if (text.trim()) out.push({ role: "assistant", content: text });
+			if (hasSubstantiveContent(text)) out.push({ role: "assistant", content: text });
 		}
 	}
 	return out;

package/src/hindsight/content.ts

@@ -43,6 +43,22 @@ export function stripMemoryTags(content: string): string {
 		.replace(LEGACY_RELEVANT_MEMORIES_REGEX, "");
 }
 
+// At least one letter or digit means the message carries a token a retriever
+// can actually match on. Punctuation/whitespace-only strings (e.g. the lone
+// `.` some providers emit for tool-call-only or thinking-only assistant turns)
+// are dropped before retain/recall touches them — see issue #1806.
+const SUBSTANTIVE_CHAR_RE = /[\p{L}\p{N}]/u;
+
+/**
+ * True when `content` carries at least one letter or digit. Used by retain
+ * and recall paths to drop placeholder assistant turns ("." / "..." / pure
+ * whitespace) that would otherwise pollute the bank and waste tokens on
+ * embeddings with no semantic content.
+ */
+export function hasSubstantiveContent(content: string): boolean {
+	return SUBSTANTIVE_CHAR_RE.test(content);
+}
+
 /** Format recall results into a bullet list for context injection. */
 export function formatMemories(results: RecallResultLike[]): string {
 	if (results.length === 0) return "";
@@ -197,7 +213,7 @@ export function prepareRetentionTranscript(
 	const parts: string[] = [];
 	for (const msg of targetMessages) {
 		const content = stripMemoryTags(msg.content).trim();
-		if (!content) continue;
+		if (!hasSubstantiveContent(content)) continue;
 		parts.push(`[role: ${msg.role}]\n${content}\n[${msg.role}:end]`);
 	}
 

package/src/hindsight/transcript.ts

@@ -9,7 +9,7 @@
 
 import type { AssistantMessage } from "@oh-my-pi/pi-ai";
 import type { SessionEntry } from "../session/session-entries";
-import type { HindsightMessage } from "./content";
+import { type HindsightMessage, hasSubstantiveContent } from "./content";
 
 export interface ReadonlySessionManagerLike {
 	getEntries(): SessionEntry[];
@@ -39,7 +39,7 @@ export function extractMessages(sessionManager: ReadonlySessionManagerLike): Hin
 		if (role !== "user" && role !== "assistant") continue;
 
 		const text = role === "user" ? extractUserText(msg) : extractAssistantText(msg as AssistantMessage);
-		if (text.length === 0) continue;
+		if (!hasSubstantiveContent(text)) continue;
 		messages.push({ role, content: text });
 	}