fix(agents): skip thinking tags in code spans

This commit is contained in:
Peter Steinberger
2026-01-15 09:23:10 +00:00
parent aac5b4673f
commit 7e1e7ba2d8
7 changed files with 275 additions and 6 deletions

View File

@@ -1,6 +1,8 @@
import { parseReplyDirectives } from "../auto-reply/reply/reply-directives.js";
import { formatToolAggregate } from "../auto-reply/tool-meta.js";
import { createSubsystemLogger } from "../logging.js";
import type { InlineCodeState } from "../markdown/code-spans.js";
import { buildCodeSpanIndex, createInlineCodeState } from "../markdown/code-spans.js";
import { EmbeddedBlockChunker } from "./pi-embedded-block-chunker.js";
import {
isMessagingToolDuplicateNormalized,
@@ -38,7 +40,7 @@ export function subscribeEmbeddedPiSession(params: SubscribeEmbeddedPiSessionPar
deltaBuffer: "",
blockBuffer: "",
// Track if a streamed chunk opened a <think> block (stateful across chunks).
blockState: { thinking: false, final: false },
blockState: { thinking: false, final: false, inlineCode: createInlineCodeState() },
lastStreamedAssistant: undefined,
lastStreamedReasoning: undefined,
lastBlockReplyText: undefined,
@@ -72,6 +74,7 @@ export function subscribeEmbeddedPiSession(params: SubscribeEmbeddedPiSessionPar
blockChunker?.reset();
state.blockState.thinking = false;
state.blockState.final = false;
state.blockState.inlineCode = createInlineCodeState();
state.lastStreamedAssistant = undefined;
state.lastBlockReplyText = undefined;
state.lastStreamedReasoning = undefined;
@@ -185,9 +188,15 @@ export function subscribeEmbeddedPiSession(params: SubscribeEmbeddedPiSessionPar
}
};
const stripBlockTags = (text: string, state: { thinking: boolean; final: boolean }): string => {
const stripBlockTags = (
text: string,
state: { thinking: boolean; final: boolean; inlineCode?: InlineCodeState },
): string => {
if (!text) return text;
const inlineStateStart = state.inlineCode ?? createInlineCodeState();
const codeSpans = buildCodeSpanIndex(text, inlineStateStart);
// 1. Handle <think> blocks (stateful, strip content inside)
let processed = "";
THINKING_TAG_SCAN_RE.lastIndex = 0;
@@ -195,6 +204,7 @@ export function subscribeEmbeddedPiSession(params: SubscribeEmbeddedPiSessionPar
let inThinking = state.thinking;
for (const match of text.matchAll(THINKING_TAG_SCAN_RE)) {
const idx = match.index ?? 0;
if (codeSpans.isInside(idx)) continue;
if (!inThinking) {
processed += text.slice(lastIndex, idx);
}
@@ -211,9 +221,11 @@ export function subscribeEmbeddedPiSession(params: SubscribeEmbeddedPiSessionPar
// If enforcement is disabled, we still strip the tags themselves to prevent
// hallucinations (e.g. Minimax copying the style) from leaking, but we
// do not enforce buffering/extraction logic.
const finalCodeSpans = buildCodeSpanIndex(processed, inlineStateStart);
if (!params.enforceFinalTag) {
state.inlineCode = finalCodeSpans.inlineState;
FINAL_TAG_SCAN_RE.lastIndex = 0;
return processed.replace(FINAL_TAG_SCAN_RE, "");
return stripTagsOutsideCodeSpans(processed, FINAL_TAG_SCAN_RE, finalCodeSpans.isInside);
}
// If enforcement is enabled, only return text that appeared inside a <final> block.
@@ -225,6 +237,7 @@ export function subscribeEmbeddedPiSession(params: SubscribeEmbeddedPiSessionPar
for (const match of processed.matchAll(FINAL_TAG_SCAN_RE)) {
const idx = match.index ?? 0;
if (finalCodeSpans.isInside(idx)) continue;
const isClose = match[1] === "/";
if (!inFinal && !isClose) {
@@ -254,7 +267,27 @@ export function subscribeEmbeddedPiSession(params: SubscribeEmbeddedPiSessionPar
// Hardened Cleanup: Remove any remaining <final> tags that might have been
// missed (e.g. nested tags or hallucinations) to prevent leakage.
return result.replace(FINAL_TAG_SCAN_RE, "");
const resultCodeSpans = buildCodeSpanIndex(result, inlineStateStart);
state.inlineCode = resultCodeSpans.inlineState;
return stripTagsOutsideCodeSpans(result, FINAL_TAG_SCAN_RE, resultCodeSpans.isInside);
};
const stripTagsOutsideCodeSpans = (
text: string,
pattern: RegExp,
isInside: (index: number) => boolean,
) => {
let output = "";
let lastIndex = 0;
pattern.lastIndex = 0;
for (const match of text.matchAll(pattern)) {
const idx = match.index ?? 0;
if (isInside(idx)) continue;
output += text.slice(lastIndex, idx);
lastIndex = idx + match[0].length;
}
output += text.slice(lastIndex);
return output;
};
const emitBlockChunk = (text: string) => {