Clickbaits are deceptive headlines or social media posts designed to attract clicks, often for profit or commercial gain. The spread of clickbait significantly impacts users negatively, misleading them and putting them at risk of click-jacking attacks. Different from fake news, the crucial problem in clickbait detection is determining whether the headline matches the corresponding content. Most existing methods for detecting clickbait rely on computing semantic similarity between headlines and content, but the significant differences in length and semantic features make this approach challenging. To address this problem, we propose a prompt-tuning method for clickbait detection via text summarization. In this paper, text summarization is introduced to summarize the contents, and clickbait detection is performed based on the similarity between the generated summary and the contents. Specifically, we first introduce a two-stage text summarization model, leveraging pre-trained language models to produce high-quality news summaries. Then, both the headlines and the generated summaries are used as inputs for prompt-tuning. Additionally, five strategies are employed to integrate external knowledge to improve the performance of clickbait detection. Extensive experiments on well-established clickbait detection datasets demonstrate that our method achieves state-of-the-art performance, exceeding the latest SOTA method by 4.42% points in accuracy and significantly enhancing Precision, Recall, and F1 scores.