This Anaconda MCP Beta Agreement (“Agreement”) is between Anaconda, Inc. (“Anaconda”) and the individual identified at acceptance below (“Participant”). This Agreement supplements and is governed by Anaconda’s Terms of Service at anaconda.com/legal/terms/terms-of-service (“Terms”). Capitalized terms not defined here have the meanings in the Terms. This Agreement is effective upon Participant’s clickthrough acceptance.
A “Beta Offering” means the Anaconda product, feature, or functionality described in Exhibit A to this Agreement, Anaconda MCP, made available on a pre-release or evaluation basis. Anaconda will identify the Beta Offering by labeling it “Beta,” or a similar designation within the Platform or in documentation.
1. Your Obligations
What you agree to do when using the Beta Offering.
You agree to: (a) use the Beta Offering lawfully and per the Terms and Anaconda’s Acceptable Use Policy; (b) not use it in production or mission-critical environments; (c) provide Anaconda with reasonable information about your use of the Beta Offering upon request, including bug reports and general usage feedback through the channel identified in Exhibit A; and (d) comply with any confidentiality obligations that apply under Section 2.
2. Confidentiality
The Beta Offering may or may not be confidential. Exhibit A will tell you which.
If designated as confidential in Exhibit A, the Beta Offering and any non-public features, functionality, or related information (“Beta Confidential Information”) are subject to the following: Participant agrees to keep Beta Confidential Information strictly confidential, use it solely to evaluate the Beta Offering, and not disclose it to any third party without Anaconda’s prior written consent. Participant must report suspected security vulnerabilities confidentially through the channel identified in Exhibit A before any public disclosure. This obligation survives expiration of this Agreement for one (1) year. Standard exceptions under Section 17 of the Terms apply.
If confidentiality is not designated in Exhibit A, no confidentiality obligation applies to the Beta Offering. However, if Participant intends to publicly disclose or discuss any suspected security vulnerability or bug in the Beta Offering, Participant must first report it directly to Anaconda through the channel identified in Exhibit A at least forty-eight (48) hours prior to such disclosure.
3. As-Is; No Support
The Beta Offering comes with no guarantees, no support, and no SLAs.
ALL BETA OFFERINGS ARE PROVIDED “AS IS” AND “AS AVAILABLE” FOR EVALUATION PURPOSES ONLY. ANACONDA MAKES NO WARRANTIES OF ANY KIND. THERE IS NO SUPPORT, NO SLA, NO UPTIME COMMITMENT, AND NO SERVICE CREDITS. ANACONDA DOES NOT REPRESENT THAT THE BETA OFFERING WILL EVER BE MADE COMMERCIALLY AVAILABLE. YOU USE THE BETA OFFERING AT YOUR OWN RISK, AND YOU ARE SOLELY RESPONSIBLE FOR ANY DATA YOU SUBMIT. ANACONDA HAS NO OBLIGATION TO STORE, MAINTAIN, MIGRATE, OR RETURN ANY DATA YOU SUBMIT TO THE BETA OFFERING, AND WILL NOT PROVIDE DATA PORTABILITY OR TRANSFER UPON TERMINATION OR DISCONTINUATION OF THE BETA OFFERING. ANACONDA MAY MODIFY, SUSPEND, OR DISCONTINUE THE BETA OFFERING AT ANY TIME WITHOUT LIABILITY.
4. Liability
Anaconda’s liability for the Beta Offering is limited.
Anaconda’s liability is limited as set forth in Section 19 of the Terms. For Beta Offerings provided at no charge, Anaconda’s aggregate liability shall not exceed USD $100. Where Exhibit A specifies fees, the liability cap in Section 19 of the Terms applies.
5. Term & Termination
This Agreement continues until ended. Anaconda can end your access at any time.
This Agreement remains in effect until terminated by either party on written notice, or upon expiration of the Beta Period set forth in Exhibit A. Anaconda may terminate or suspend your access to the Beta Offering or this Agreement at any time, for any reason, with or without notice. Upon termination, you must cease use and destroy any applicable Beta Confidential Information. Sections 2, 3, 4, and 7 survive termination.
6. Fees
The Beta Offering is free during the Beta Period unless Exhibit A says otherwise.
Unless Exhibit A specifies otherwise, the Beta Offering is provided at no charge during the Beta Period. This does not grant free access to any underlying products or dependencies that otherwise require a paid license.
7. Eligibility
Who can participate.
This Agreement is open to users accessing the Anaconda Platform under the Terms. By accepting, you confirm that: (a) you are currently using the Platform under the Terms; (b) you qualify for free use under Section 1(a) of the Terms, or hold a valid paid Subscription; (c) you are not using the Beta Offering as a substitute for a paid license you would otherwise be required to obtain; and (d) if scope-limited per Exhibit A, you meet the geographic, organizational, or other eligibility criteria specified therein. Anaconda may verify eligibility and revoke access at any time.
8. Feedback & IP
Anaconda owns its platform. You own your content. Feedback you choose to share belongs to Anaconda.
Anaconda retains all rights to the Beta Offering. You retain ownership of Your Content. Any Feedback you voluntarily provide is licensed to Anaconda as set forth in Section 16(c) of the Terms. You have no obligation to provide Feedback beyond what is specified in Section 1(c) above.
Acceptance
By clicking on the “agree” (or similar button or checkbox) that is presented to you before using, downloading, installing or accessing the Beta Offering, you confirm you are bound by these Terms.
EXHIBIT A — ANACONDA MCP BETA OFFERING DETAILS
This Exhibit A is part of and incorporated into the Anaconda MCP Beta Agreement. The terms of the Agreement apply to the Beta Offering described below.
| Beta Offering Name | Anaconda MCP |
| Beta Offering Description | The Anaconda MCP is a unifying Model Context Protocol (MCP) server that bridges the gap between AI coding assistants and the Anaconda ecosystem. |
| Anaconda Beta Contact | https://anaconda.canny.io/anaconda-mcp-beta |
| Beta Period | Begins 27 MAY 2026; terminable by Anaconda at any time |
| Eligibility | No changes — Agreement defaults apply (global). |
| Confidentiality | Public beta. No confidentiality obligation applies beyond Section 2 security reporting. |
| Fees | No charge during Beta Period. |
| Participant Obligations | Agreement Section 1 defaults apply. In addition: Feedback: Provide monthly written feedback via https://anaconda.canny.io/anaconda-mcp-beta Bug Reporting Channel: https://anaconda.canny.io/anaconda-mcp-beta |
| Personal Data | Participant will not submit personal data to Anaconda unless Anaconda has agreed in writing to accept and process such data from Participant. |
| Product-Specific Terms | 1. Scope; Evolving Capabilities. These Product-Specific Terms apply to the MCP as it exists at the time of your acceptance and as it may develop over time. Anaconda intends to expand the capabilities of the MCP through new tools, data sources, functions, and integrations, and these terms are intended to govern the MCP in its current form and to remain applicable as the Service evolves without requiring amendment. Certain provisions below describe functionality, including security assessment, compliance checking, audit trail generation, and automated environment or code modification, that may not be present in the current release of the MCP. The inclusion of a term addressing specific functionality is not a representation, commitment, or guarantee that such functionality will be developed, released, or made available. Anaconda expressly disclaims any obligation to deliver any capability described herein that is not currently available. New capabilities added to the MCP are governed by this Agreement unless a specific capability is accompanied by a separate written agreement. Your continued use of the MCP following the addition of new capabilities constitutes your acceptance of these terms as applied to those capabilities. If you do not wish to be bound by these terms as applied to a new capability, you must cease use of that capability and, if applicable, the MCP. 2. Execution Environment. Anaconda MCP operates as a tool server that runs in the same environment as the AI agent or platform that invokes it. It inherits the access, credentials, and permissions of that environment. You are responsible for ensuring that the environment in which you deploy and run the MCP, including the permissions you grant to the AI agent invoking it, is configured appropriately for your use case. 3. Your AI Agent; Bring Your Own AI. 3.1. Agent Behavior; Permission Variability. AI agents and platforms differ in whether they seek user confirmation before executing a tool call. Some will prompt you before executing commands with significant or irreversible effects; others may execute automatically, including where permission-seeking has been disabled. Anaconda does not control this behavior. You are responsible for understanding and configuring the permission model of your agent or platform before deployment. The environment in which the MCP runs depends on the agent or platform you use. If your agent runs locally, the MCP runs on your local machine. If your agent runs in a cloud environment, the MCP runs within that environment, and your use of it is governed by that platform’s own terms. Anaconda is not responsible for the behavior, availability, security practices, data handling, costs, or outputs of third-party platforms. 3.2. Compute and Resource Usage. When the MCP is accessed through a cloud-hosted AI agent or platform, that agent or platform may incur compute costs, API usage fees, or other resource consumption charges. Anaconda is not responsible for any such costs or charges. You are responsible for configuring, monitoring, and controlling the resource usage of any AI agent or platform you use in connection with the MCP. 3.3. Third-Party AI Agents and Platforms. The MCP is designed to be compatible with multiple AI agents and platforms. Anaconda does not endorse any specific AI agent or platform, and is not responsible for the behavior, availability, security practices, data handling, or outputs of any third-party agent or platform. Your use of any third-party agent or platform is governed by that party’s own terms. Data processed by third-party agents or platforms in connection with the MCP is subject to those parties’ own terms and privacy policies, and Anaconda’s data handling commitments apply only to data processed directly by Anaconda’s infrastructure. 3.4. Your Responsibility; Oversight. You are responsible for all actions taken by AI agents you configure, direct, or permit to access the MCP, whether or not specifically anticipated, including conda commands, package or environment modifications, automated repairs, and code changes performed by or at the direction of an agent. You acknowledge that agents using the MCP may take actions that are difficult or impossible to reverse, and that Anaconda does not guarantee prevention of unintended outcomes. You agree to apply human review appropriate to the consequences of the actions agents will take, particularly before permitting actions that are irreversible, affect production environments, modify code, are financially material, are legally binding, are externally communicative, or affect third-party systems. Agents may be manipulated by adversarial inputs embedded in content they access through the MCP, including content from community forums, documentation, or user-uploaded materials. You are responsible for assessing and mitigating that risk. 3.5. Automated and Agentic Actions. In current or future versions, the MCP may support or facilitate actions performed by AI agents on your behalf, including operations affecting your conda environments, packages, files, dependencies, and code (“Automated Actions”). Anaconda does not warrant that any Automated Action will produce a correct, complete, or intended result. Anaconda is not liable for data loss, configuration changes, code modifications, or other consequences resulting from Automated Actions, including actions that may be irreversible. 4. Informational Outputs; No Warranty on Accuracy. In current or future versions, the MCP may return informational outputs, including environment data, package and dependency information, search results, security-related information, compliance or policy assessments, and audit or provenance records. Outputs may be derived from third-party sources, including package repositories, vulnerability databases, community forums, and content uploaded by users to anaconda.org. All informational outputs are provided for informational purposes only. Anaconda does not warrant the accuracy, completeness, currency, or fitness for any particular purpose of any output of the MCP, including outputs derived from third-party or user-generated sources that have not been reviewed or endorsed by Anaconda. You are responsible for independently verifying any output before relying on it. 5. Your Outputs; Data Use. As between Anaconda and Participant, Participant owns the outputs returned to Participant through the MCP. Anaconda does not claim ownership of such outputs. Anaconda may return the same or similar outputs to other participants. Anaconda does not use the content of prompts, queries, or outputs to train, fine-tune, or improve AI models. When you use the MCP through a third-party AI agent or platform, that platform’s own data handling and model training policies apply, and Anaconda is not responsible for those policies. 6. Security Information. In current or future versions, the MCP may provide security-related information about packages, environments, or dependencies, derived from publicly available vulnerability databases and third-party security data sources. The following limitations apply to any such outputs. Security-related outputs do not constitute a security audit, penetration test, professional security assessment, or determination that any package, environment, or configuration is safe or appropriate for use, and should not be relied upon as a substitute for any of the foregoing. Security information reflects data available at the time of query and may not reflect vulnerabilities disclosed after that time. Anaconda does not warrant that all known or unknown vulnerabilities affecting any package, environment, or dependency will be detected or reported. 7. Compliance and Policy Functions. In current or future versions, the MCP may provide functionality that checks package configurations, environments, or dependencies against policy rules. Compliance and policy outputs are provided for informational purposes only. They do not constitute legal, regulatory, or compliance determinations, do not represent that you or your organization is in compliance with any applicable law, regulation, or industry standard, and should not be relied upon as satisfying any legal, regulatory, or certification requirement or as a substitute for independent legal or compliance advice. Where the MCP allows you or your organization to configure policy rules, you represent that those rules accurately reflect your compliance requirements. Anaconda is not responsible for the accuracy or legal sufficiency of customer-defined policies, or for outcomes resulting from their enforcement. 8. Audit Trails and Record Functions. In current or future versions, the MCP may generate logs, audit trails, activity records, or provenance data reflecting tool invocations, agent actions, or recommendations made during a session (“Records”). Records are provided as a convenience feature. They are not guaranteed to be complete, accurate, unaltered, or tamper-evident, and are not suitable for use in legal, regulatory, audit, or evidentiary proceedings. Anaconda makes no representation that Records meet the evidentiary or record-keeping requirements of any legal, regulatory, or certification framework. 9. Authentication. Use of the MCP requires authentication with your Anaconda account. You are responsible for the security of your authentication credentials and for all actions taken under your authenticated session. 10. Telemetry. The MCP collects usage data, including tool invocation frequency and feature usage patterns. Anaconda does not collect the content of prompts or queries submitted to AI agents in connection with the Service. For more information about Anaconda’s data practices, see Anaconda’s Privacy Policy. 11. Restrictions. In addition to the Acceptable Use Policy, you will not: (a) expose the MCP to personal data such as Social Security numbers or other government identifiers, financial account or payment card information, protected health information, biometric data, precise geolocation, children’s data, or other sensitive categories of personal data under applicable law; (b) use the MCP, or rely on its outputs, in regulated domains as a substitute for professional legal, medical, financial, or compliance advice; (c) deploy the MCP for use cases where errors could cause material harm to persons or property, irreversible loss of data, or regulatory violations, including safety-critical situations; (d) configure or permit AI agents to use the MCP to take actions on third-party systems, accounts, or services without authorization from the owner of those systems; or (e) use security, compliance, or audit outputs of the MCP as the sole basis for a security certification, regulatory filing, legal proceeding, or compliance attestation. 12. Distribution. The MCP may be distributed through various channels, including Anaconda’s own applications and third-party registries. Your use of the MCP is governed by this Agreement regardless of the channel through which you obtained or installed it. |