fix: raise OpenClaw report maxTokens to 8192 to prevent mid-sentence truncation

[litang9]

Problem

The OpenClaw deep-dive report gets truncated mid-sentence on high-activity days (e.g. #195):

今日 OpenClaw 项目展现出极高的工程迭代效率与社区活跃度。过去

Sentence cuts off at "过去"; the comparison section even notes "由于今日 OpenClaw 的原始数据源在提取时意外截断".

Root cause

OpenClaw consumes the largest input (issues+prs+releases, often 900+ items) but calls callLlm with the **default `maxTokens=4096`` — smaller than every other report:

Report	maxTokens	Input
trending	6144	small
web	8192	medium
OpenClaw	4096 (default)	largest

Output hits finish_reason=length mid-report and the truncated text is written to the issue.

Fix

Pass 8192 explicitly (src/index.ts), matching the web report:

- return await callLlm(buildPeerPrompt(cfg, issues, prs, releases, dateStr, 50, 30, lang));
+ return await callLlm(buildPeerPrompt(cfg, issues, prs, releases, dateStr, 50, 30, lang), 8192);

Co-Authored-By: Claude noreply@anthropic.com