Kimi Reaches Settlement with Open Source Community: Risk Control Update Mistakenly Affects Third-Party Tool, "Double Standard PR" a Misunderstanding

According to Dongcha Beating monitoring, the Kimi Code team under the Moon's Dark Side reached a settlement with the open-source community on the evening of May 26, completely quelling the subscription risk control and PR double standard crisis triggered by the ban incident. Open-source developer Leechael announced the removal of previous inflammatory remarks and confirmed that the pi-provider-kimi-code plugin will continue to be maintained. Future paid subscription members can safely continue to use the Kimi Code Plan in third-party tools such as Pi Coding Agent.

Kimi Code team member Young issued an official Chinese statement explaining the technical details of the ban defense. The team recently upgraded the risk control defense strategy to combat relay station sales and abnormal requests, using the User-Agent identifier as one of the criteria. However, the risk control model did not adequately consider the existence of many third-party Code Agents (such as Pi) designed as Kimi CLI User-Agents, causing the system to incorrectly classify normal requests as malicious spoofing, unintentionally harming a large number of compliant subscription users.

The double standard controversy was actually a misunderstanding caused by a communication gap across multiple channels. Young explained to this channel that since most domestic large-scale model developers are mainly concentrated in the Moon's Dark Side's official Feishu group, the team provided explanations and reassurance within Feishu immediately after the incident. However, English responses on Platform X were delayed due to time zone differences and cross-platform operations, inadvertently creating a double standard perception of only apologizing to the external network. Currently, the Moon's Dark Side has comprehensively adjusted its risk control strategy, all restricted accounts have been restored to normal, and they have begun proactively contacting third-party developers to optimize User-Agent identification.

Following the settlement, Leechael revealed that Kimi developers have had in-depth communication with Leechael, reestablishing basic confidence in the domestic large model ecosystem, and have begun writing a dedicated prefix caching optimization guide for Kimi's personal intelligent agent.