1. Product and scope
iAgent is a local desktop app focused on workflow assistance in its current public scope. It helps users turn plain-language requests into reviewable workflows, but it does not replace professional judgment, business rules, legal advice, or internal approval controls.
Current public installers cover macOS and Windows. Linux, team collaboration, full organization management, and deep external-system integration belong to the public service scope only after the Download page, release notes, or official docs clearly say they are live.
Marketing copy is not a substitute for product facts. Functional scope follows backend behavior, public docs, release notes, and the actual application interface.
2. Account registration and AI interaction logs
By registering, signing in, purchasing, or using iAgent, users agree to these terms and the privacy policy. Users who do not agree should not register or continue using the product.
iAgent cloud AI features may send user prompts, document content placed into model context, table content, converted file text, tool call parameters, model responses, model name, token usage, cost, latency, status code, and error information to the iAgent cloud proxy, model providers, and necessary cloud service providers.
iAgent may use Cloudflare AI Gateway or similar cloud logging systems to store AI interaction logs for debugging, quality analysis, security review, abnormal request tracing, cost control, billing, and service operations. Users should avoid submitting data they are not allowed to process, secrets, payment information, identity numbers, complete sensitive contracts, or other content they do not want to enter the cloud model path.
3. User responsibilities
Users are responsible for making sure imported, processed, and exported data is legally sourced and that they have permission to process it. Users must not use iAgent to process data they are not allowed to access or to bypass internal controls.
Users must review output from models, rules, scripts, and workflows. High-risk actions such as external-system writes, regulated filing, audit support, statutory disclosure, system import, and file overwrite require human confirmation.
Users may not use iAgent for unauthorized access, data leakage, malicious automation, unlawful scraping, regulatory evasion, or other unlawful activity.
4. Generated content and accuracy
iAgent may generate classification suggestions, rule text, script text, report notes, anomaly warnings, or other assistive content. This content improves workflow speed but should not be treated as automatic business, legal, or audit conclusions.
For amount handling, precision rules, record balance, report scope, and file writes, the product should prefer deterministic code and reviewable rules. Users still need to confirm results against source material and business policy.
If generated content conflicts with original documents, company policy, accounting rules, tax requirements, or human review, those sources take priority.
5. Refund policy
Users who purchase a paid subscription, managed model credits, account credits, or other metered services through the official site or an iAgent-designated payment channel may request a refund within 14 days from the payment date. iAgent will review the order, account, and usage records to calculate any refundable amount.
The refundable amount is calculated from the amount paid minus usage already consumed. Managed model credits, account credits, metered service charges, and third-party services already delivered are not refundable.
For orders completed through App Store, Google Play, Lemon Squeezy, or another third-party channel, the refund request flow, eligibility review, and processing time follow that channel's rules. iAgent may help verify order and usage records when needed.
iAgent may refuse a refund request or stop related services in cases involving unlawful use, service abuse, fraudulent transactions, malicious refund activity, continued use of service benefits after refund, or violation of these terms.
6. Updates, support, and changes
Desktop versions ship through the public update channel. Platform builds, signing, and upload steps may finish at different times, so public releases may not be synchronized across all platforms.
To keep core features such as document parsing, spreadsheet handling, OCR, and automated workflows working properly, the iAgent desktop app may automatically download, install, or update required local runtime components after installation, after first launch, or when a related feature is first used. These components may include the Python Runtime, document-processing components, and security update components. They are installed in the local app data directory and verified through version metadata, hashes, and signatures.
Launch-stage support boundaries follow the Download page, release notes, and public docs. Capabilities not listed there should not be understood as public commitments.
These terms may change as the product and release process evolve. Formal terms updates should be published on the official site with an effective date or applicable version.
7. Review notice
This page describes the current public service scope and does not replace professional judgment by counsel or a compliance owner. Liability limits, dispute resolution, governing law, and cross-border data language require separate confirmation.
Before public adoption, the product owner should verify product facts and a legal or compliance reviewer should verify whether the language fits the target market.
If these terms conflict with an order, invoice, privacy policy, or applicable law, the applicable law and more specific written agreement control.