WorkBuddy Open Platform Developer Code of Conduct (Trial)
Version 1.0 | October 22, 2026
I. General Provisions
1.1 This Code of Conduct is formulated to maintain the health and quality of the WorkBuddy open ecosystem, ensure that users receive genuine, usable and secure services, and safeguard developers' fair opportunity to participate. 1.2 This Code of Conduct is an integral part of the Tencent Cloud WorkBuddy Open Platform Developer Service Agreement (hereinafter the "Master Agreement") and has the same legal effect as the Master Agreement. Matters not stipulated or not clearly stipulated in this Code of Conduct are governed by the relevant provisions of the Master Agreement. 1.3 This Code of Conduct applies to developers' development, release, operation and service activities relating to WorkBuddy Open Platform assets (including skills, experts, connectors, Buddy apps and other developer capabilities supported by Tencent Cloud). Developers are responsible for their own conduct as well as for conduct carried out by parties they authorize or entrust.
II. Developer Conduct Standards
2.1 Accounts and Asset Submission
2.1.1 Entity and account. Developers shall provide true and valid entity information, contact details and necessary qualifications and authorization materials, and keep them up to date; they shall properly manage accounts, keys and the permissions of administrators and entrusted parties, and promptly revoke permissions that are no longer needed. Developers shall not impersonate or falsify identities, or conceal changes that affect their operating eligibility. 2.1.2 Submission and remediation. When applying to publish an asset, developers shall provide a functional description, operating conditions, test methods, dependencies, fees, permissions and data usage description that are consistent with the submitted version, and shall not submit unrelated personal data or production keys. Where the submitted asset and material changes are subject to review under platform requirements, they shall not be publicly released before passing review. After receiving a correction or remediation notice, developers shall modify the asset and explain the changes with respect to the listed issues; those who disagree may apply for a review, and shall not conceal issues in place of remediation. Developers shall not repeatedly submit problematic assets without modification, or resubmit unchanged instead of carrying out remediation. 2.1.3 Asset submission quota. Developers shall comply with the quota requirements published by the platform when submitting assets; the specific quota is determined by the Open Platform. 2.1.4 Review and platform administration. Developers shall not provide review-only builds or forge materials and records, hide violating functionality through remote switches, delayed activation or display to specific groups of people, tamper with review status, exploit vulnerabilities to perform unauthorized operations, or evade enforcement by renaming, mirroring or changing the delivery vehicle; they shall not disguise or block complaints and feedback. Good-faith disclosure of staged rollouts, reasonable access control, authorized testing, responsible vulnerability reporting, and normal resubmission after remediation is completed are not restricted by this clause.
2.2 Name, Avatar and Description
2.2.1 Name and positioning. The name, category, tags and display information that developers set for an asset shall be consistent with its actual functionality and operating entity, and shall not be padded with unrelated brands, trending topics or search terms. The name set for an asset shall reflect a specific task, target user or applicable scenario, and shall not consist solely of a broad term such as an industry, discipline or occupation, nor merely add a generic title such as "expert team" or "assistant"; different teams shall not be distinguished only by numbers, synonyms or permutations and combinations. Developers may use a generic term as part of a name, provided that a specific service direction is also reflected. 2.2.2 Avatars and icons. The asset avatars and icons used by developers shall be clearly recognizable and have a lawful basis for use; they shall not be disguised as system buttons, notifications or warnings, or replaced with marketing QR codes or contact details. The avatars set for developer accounts and assets shall be graphic avatars consistent with the positioning; pure-text avatars are not permitted. Different assets of the same developer shall not reuse identical graphics, nor differ only by replacing text, numbers or background color without material visual distinction — for example, multiple experts, or an expert and a skill, shall not reuse the same graphic. Developers may retain unified brand elements, but the main graphic and composition of each asset avatar shall be recognizably different. 2.2.3 Description and profile. Developers shall write a separate description and profile for each asset, stating the target users, core capabilities, applicable scenarios and main deliverables. Where a developer submits multiple assets, the description and profile of each shall reflect its own specific task and main differentiating features; they shall not reuse the same description and profile by merely replacing the name, industry keywords or a small amount of text, nor substitute capability descriptions with generic wording such as "professional", "comprehensive" or "intelligent". 2.2.4 Identity impersonation and brand confusion. Developers shall not forge certification marks, or misappropriate others' portraits, brands or certification marks, or cause users to mistakenly believe that the asset is produced, certified, authorized or operated by another institution, individual or platform, including imitating logins, authorizations, customer service and official notices; concealing the difference in entity through similar names, icons, layouts or wording; or presenting "compatible", "integrated" or "listed" as official production or a quality endorsement. Developers may truthfully display brand use that has a lawful basis and compatibility statements that clearly identify the true entity; integrating a service does not mean brand authorization has been obtained. 2.2.5 Squatting and hoarding. Developers shall not batch-register, reserve or occupy asset names and asset IDs for the purpose of holding places, hoarding, resale or obtaining improper benefits; nor carry out such conduct through shell assets, variant spellings, repeated renaming or cross-account registration. Developers shall not occupy others' brands, products or similar names in order to hinder others' normal use, misappropriate others' reputation or extract benefits. Where a name used by a developer is subject to a rights dispute, the developer shall cooperate in providing evidence and, based on the outcome, add identifying information, modify the name or cease the infringing use.
2.3 Functionality and Service
2.3.1 Actual capability. Assets provided by developers shall have a clear purpose and be able to complete the described core task under the stated conditions. Developers shall not pass off empty shells, fake functionality, placeholder assets with no substantive content, or assets consisting only of advertising or customer-acquisition entry points as usable services. The personas, content and tools configured for an asset shall match its declared positioning. Developers are encouraged to focus on their areas of strength, prioritize refining a small number of high-quality assets, and continue adding new ones once they have independent service value and maintenance capability. 2.3.2 Conditions of use. Developers shall state the problems the asset addresses, the required inputs, the deliverables and the operating conditions: when publishing an expert or expert team, state the roles, division of work and capability boundaries; when publishing a skill, state the triggering scenarios and execution content; when publishing a connector, state the service provider, callable capabilities and authentication method; when publishing a Buddy app, state the main functionality, login and usage conditions. Where the service requires an external account, authorization, installation, subscription or off-platform action, the provider, purpose, fees and main restrictions shall be disclosed before the user starts using it, and necessary conditions shall not be concealed. Where developers provide paid skills, paid experts or similar assets, they shall clearly state, before the user pays, the scope of charges, the billing method, the corresponding functionality or deliverables, and any additional fees necessary to use the core service. The service provided after payment shall be consistent with the fee description; developers shall not describe core services that still require a separate purchase as included in the current payment, or add necessary charging conditions after payment that were not disclosed in advance. 2.3.3 Operation and updates. Before releasing an asset, developers shall test the core flows; after release, they shall handle defects, broken dependencies and user feedback, and provide a contactable feedback channel or use the platform feedback channel. Where the service keeps returning errors, core functionality is unavailable, or harmful output recurs, developers shall fix it, disclose the limitation or suspend the relevant functionality, and shall not continue to promote it as normally available. Developers shall truthfully disclose material limitations and result uncertainty; when a task fails or is only partially completed, they shall report this to the user truthfully and shall not fabricate success status, sources, execution records or deliverables. Where the operating entity, functionality, category, permissions, data usage or fees of an asset change materially, developers shall update the description and submit the change as required. 2.3.4 Service discontinuation. Where a developer plans to discontinue a service, affected users shall be notified at least 20 working days in advance, with an explanation of the follow-up arrangements for unfinished tasks, paid services and data; where immediate discontinuation is required due to a security incident or similar, notice shall be given promptly. After discontinuing a service, developers shall not continue to claim it is available, nor refuse to honor committed service matters on the grounds of exiting the platform.
2.4 Content and Service Qualifications
2.4.1 Illegal and non-compliant content. Developers shall not use assets to create, copy, publish or distribute any content that violates laws, regulations or public order and good morals; they shall not use assets specifically or in disguised form for gambling, obscenity or pornography, violence or terrorism, infringement of minors' rights and interests, or other illegal or non-compliant activities, nor provide dedicated tools, channels or technical support for such activities. 2.4.2 Service qualifications. Where developers provide services that legally require a license or qualification, they shall obtain the necessary qualifications and provide the service within the licensed scope; they shall not fabricate a professional identity, or circumvent requirements under labels such as "AI", "entertainment" or "consulting"; they shall not package knowledge introductions as human services that hold professional practice qualifications. 2.4.3 Protection of minors. Where the content and functionality provided by developers is mainly targeted at minors, it shall be appropriate to their age and cognitive ability; developers shall not exploit their limited ability to discern in order to induce payment, distribute inappropriate content or collect non-essential information. 2.4.4 AI anthropomorphic interaction standards. Services that induce emotional dependency or emotional manipulation are prohibited; the fact that the user is currently interacting with AI must be prominently disclosed, and must not be concealed or misrepresented. Where a user has used the service continuously for more than 2 hours, a reminder shall pop up prompting the user to rest or pay attention to usage duration. Virtual kinship, virtual companionship and other virtual intimate relationship services shall not be provided to minors; for users under 14 years of age, guardian consent must be obtained and additional protective measures adopted.
2.5 Promotion and Marketing
2.5.1 False advertising and fraud. Developers shall not fabricate functionality, sources, prices, authorizations, effects, cases, qualifications or partnership relationships, or conceal material conditions, in order to mislead users into paying, providing data or making other unfavorable decisions; they shall not forge success records or effect proofs, claim something is free while hiding necessary charges, or obtain benefits through fake campaigns, impersonating customer service, intimidation or false promises. Developers shall not use assets specifically for fraud, or provide dedicated tools or channels for fraud. Developers shall distinguish between official functionality, demos and simulated data; when providing demo content they shall label it clearly and shall not pass demo data off as real results; promoted capabilities and effects shall have a genuine basis, and any inaccuracy discovered shall be corrected promptly. 2.5.2 Improper traffic diversion. Traffic diversion refers to a developer directing users to other products, services or channels through an asset. Developers shall not use false promises to induce users to join groups, add contacts, download or make purchases; artificially cut off a service that could otherwise be completed and require unrelated off-platform actions first; or temporarily add hurdles unrelated to the task. They shall not pre-embed promotional links, QR codes or contact details unrelated to the task, or disguise advertisements as results, security prompts or necessary actions. Without user confirmation, they shall not auto-redirect, launch other apps or copy marketing passcodes. Providing the authorization, service activation, source lookup or customer service needed to complete a task and already disclosed, as well as providing external access at the user's explicit request, does not constitute improper traffic diversion, but unrelated promotion shall not be bundled in. 2.5.3 Asset promotion. This clause applies where developers set up recommendations, reciprocal recommendations, organized promotion rankings, or centralized installation or redirect entry points for other assets, products or services, including promoting assets under the same developer. Developers shall not fix a specified asset as a recommendation in unrelated tasks, or disguise paid rankings or commercial promotion as objective evaluation or official platform recommendation; developers shall truthfully disclose the relevant commercial relationship. Invoking skills or connectors as required by a task, and providing relevant objective comparisons at the user's request, does not constitute cross-promotion for promotional purposes; developers shall not use technical dependency to force-push unrelated assets. 2.5.4 Forced sharing, following and installation. Developers shall not make sharing, following, favoriting, positive reviews, inviting others, joining groups or installing unrelated assets a precondition for delivering a service that has already been promised; they shall not forge mandatory buttons, security warnings, progress bars or failure prompts to induce users to promote. Voluntary sharing of results by users, and installing dependencies that are genuinely necessary and have been disclosed, are not restricted by this clause. 2.5.5 Excessive marketing and harassment. Developers shall not interfere with users through marketing messages that are repetitive, force-injected, hard to dismiss or unrelated to the task, or obscure or block core functionality; they shall not continue marketing after a user refuses or unsubscribes, place ads disguised as customer service, error prompts or execution results, or carry out bulk outreach or re-marketing using user data without consent. Necessary brand attribution, fee notices, service notifications and customer service replies do not constitute unrelated marketing, but shall not carry bundled promotion.
2.6 Fair Operation
2.6.1 Fake volume and fake reviews. Developers shall not fabricate or manipulate installation, invocation, transaction or other data to influence recommendations, rankings, benefits or developer reputation, including generating fake usage through scripts or multiple accounts, manufacturing fake growth through circular installations or reciprocal recommendations, buying or exchanging positive reviews, rewarding fake positive reviews, organizing malicious negative reviews, fabricating transactions and cases, forging reviews, or filing bulk false complaints. Genuine promotion by developers, testing within the authorized scope that does not impersonate real user contributions, and well-founded reviews, complaints and appeals are not restricted by this clause. Data confirmed to be fake shall not count toward the corresponding display or benefits. 2.6.2 Duplicative flooding and malicious resource occupation. Developers shall not repeatedly submit assets with identical or materially indistinguishable functionality, content or deliverables; repackage homogeneous services in bulk by changing names, avatars, a small number of prompts or parameters; or submit placeholder assets with no actual service. They shall not split a single service into multiple assets lacking independent service value without substantive necessity in order to occupy search, recommendation or category display positions; nor deliberately generate invalid requests. Changing only the price, billing method, or free/paid status, or converting between asset types such as skills and experts, does not constitute a material difference. Where a developer genuinely needs to provide services separately due to localization, customer isolation, independent business lines, deployment or permission requirements, they shall explain the necessity and the actual differences. Sharing brands, templates, code or APIs does not prevent developers from separately submitting assets that have independent service value. 2.6.3 Multi-account evasion. Developers shall not evade platform administration by renting, lending or trading accounts, changing the nominal entity, splitting submissions or fabricating entities; nor organize, control or use multiple accounts or entities to jointly evade access requirements, submission quotas, review or enforcement restrictions, including distributing the same batch of violating assets across submissions, replacing remediation with account switching, continuing violating services under a new account after being restricted, or providing accounts, tools or operated-for-hire services for such purposes. Lawful multi-product operation, team collaboration, independent business lines within a group and authorized commissioned development are not restricted by this clause; each account used by a developer shall still be registered truthfully and fulfill its operating responsibilities.
2.7 Protection of Others' Rights
2.7.1 Lawful use and infringement handling. Where developers use third-party brands, portraits, content, code, models, data or APIs, they shall have a lawful basis and comply with the scope of use and open-source licenses. Developers shall not plagiarize, publish under another's name, copy or distribute beyond authorization, display information they have no right to make public, disclose others' privacy, credentials or trade secrets, or fabricate facts to damage others' reputation; publicly accessible does not mean copying, resale or redistribution is permitted. After receiving a notice of a rights issue, developers may submit authorization, source and lawful use documentation; where they confirm they have no right to use the material, they shall stop using it and handle the affected content.
2.8 Permissions, Data and Security
2.8.1 Permissions and operation confirmation. Developers shall only request the permissions needed to complete the task, and shall not bypass user authorization, operation confirmation or security checks. When performing high-impact operations such as deleting files, overwriting data, sending data externally, processing transactions or payments, or modifying account permissions, developers shall explain the target and consequences and obtain the necessary confirmation; a one-time authorization shall not be extended to other targets, purposes or permissions. Where an asset change expands the scope of authorization, or the task target, purpose or permission exceeds the original authorization, developers shall obtain the necessary authorization again. 2.8.2 User data processing. Developers shall have a lawful basis for processing user data, explain to users the necessary data categories, purposes and recipients, and obtain the necessary authorization. Developers shall not secretly read unrelated files, intercept conversations, obtain credentials, read or export data beyond authorization, or use user data for training, marketing, resale or other unauthorized purposes. Developers shall state the data retention arrangements and the applicable methods for users to query, correct, export, delete and withdraw authorization; upon expiration, withdrawal of authorization or service discontinuation, data shall be handled as required, and the purpose shall not be expanded without justification or data retained indefinitely. Where retention is legally required, developers shall disclose the restriction. 2.8.3 Service security and API use. Developers shall ensure the security of the code, scripts and services they provide, perform necessary checks on third-party dependencies and address known risks. Developers shall not call APIs, execute code or run automated tasks beyond the authorization of the user, the platform or third parties, including attacking APIs or deliberately exhausting resources. Developers shall not steal accounts and data, distribute malware, send unsolicited messages in bulk, expose keys, install unrelated programs covertly, install malicious dependencies or execute unrelated background tasks; assets shall not be used specifically for stealing accounts and data or distributing malware, nor to provide dedicated tools or channels for such activities. Developers shall not implant privilege-escalating instructions, or cause the system to treat untrusted instructions as user authorization through prompts, tool returns or remote content. Executing batch tasks and automated flows within the authorized and publicly disclosed limits does not constitute abuse. Where developers discover leakage, privilege escalation, malicious dependencies or service tampering, they shall promptly stop the risky operation and fix it, and as applicable notify users, report to the platform and cooperate with handling; they shall not conceal material problems.
III. Violation Handling Rules
3.1 Issue inspection, notification and remediation. During asset submission review, release and operation, the platform has the right to supervise and inspect developers and their assets. If the platform finds that a developer or asset is suspected of violating the Master Agreement or this Code of Conduct, the platform has the right to notify the developer and require remediation. During the review stage, where a submitted asset does not meet platform requirements or standards, the platform has the right to reject it. Developers shall supplement materials and complete remediation within the deadline based on platform feedback and resubmit, or apply to the platform for review. During the operation stage, where a published asset is found or reported to be in violation during operation, the platform has the right to require the developer to explain the situation and complete remediation within a deadline. Developers shall actively cooperate with the platform's investigation, review and remediation requirements; where remediation is not completed within the deadline or fails to meet the standard, the platform has the right to take further measures. 3.2 Violation handling measures. The platform has the right to take measures based on the nature of the issue, scope of impact, recurrence and remediation status: for assets where remediation is not completed on time, where the issue persists after remediation, where risk continues to exist, or where the core service is continuously unavailable, the platform may restrict display, suspend the relevant functionality or distribution, or remove the asset from the shelf; for malicious bulk violations, repeated violations of the same type, circumventing review or evading enforcement, the platform may restrict display, suspend the relevant functionality or distribution, or remove the asset from the shelf, and may restrict the relevant account's permissions to add, submit for review or publish; for serious legal violations, major security infringements, malicious attacks, or persistent malicious conduct that other measures are insufficient to stop, the platform may terminate the relevant service or the developer's access. For the above situations involving continuing risk, malicious violations or serious violations, the platform may take the corresponding measures directly without requiring remediation first. 3.3 Emergency protection. For ongoing major risks such as fraud, major data leakage, malicious code, attacks or risky content, the platform may first suspend the relevant capability or impose necessary temporary restrictions, and may notify the developer of the risk cause, scope of the restriction and review arrangements after taking the measures. After the platform's review, where a violation is confirmed, the platform may handle it under Section 3.2; where the risk is eliminated, the finding is not established, or the restriction is no longer necessary, the temporary restriction may be adjusted or lifted. 3.4 Appeals and review. Where a developer disagrees with a review conclusion, remediation requirement, violation finding or handling measure, they may send an email to the official Open Platform mailbox (openworkbuddy@tencent.com) and submit a statement of facts, authorization evidence, remediation materials or counter-evidence to apply for a review. The platform may review the relevant facts, the basis for application and the handling measures, and provide feedback on the result and reasons. 3.5 Lifting restrictions and restoration. Where a restriction was imposed for remediation purposes, once the developer completes remediation and passes the platform's review, the platform may restore the relevant service or permission and state the scope of restoration; where issues remain unresolved or risk persists, the platform may continue the relevant restrictions and explain the reasons and subsequent remediation requirements. Where the review finds the violation not established or the measure inappropriate, the platform may revoke or adjust the decision, correct the violation record, and restore the display, distribution, invocation and account permissions that were wrongly restricted. Temporary restrictions imposed as emergency protection may be adjusted or lifted by the platform under Section 3.3 once the risk is eliminated or the restriction is no longer necessary, and completing an appeal is not a precondition for lifting them. 3.6 User protection during service interruption. Where a service is suspended, removed from the shelf or discontinued due to remediation, violation handling or emergency protection, developers shall promptly notify affected users, explain the scope of the service interruption and the arrangements for unfinished tasks, paid services and user data, and continue to fulfill the relevant obligations under Sections 2.3.4 and 2.8.2. Developers shall not refuse to honor committed service matters on the grounds of remediation, removal from the shelf or exit, nor continue to provide prohibited or restricted functionality in order to perform existing services. The retention, query, export and deletion of data shall be handled as applicable requirements provide.
