1. What the document is: Anthropic's "Software Directory Terms" and "Software Directory Policy" — governs third-party developers submitting MCP servers, Skills, plugins, and apps to Anthropic's directories (Claude's equivalent of OpenAI's App Developer Terms / Connector ecosystem).
What this wiki found — complete, every page
Presence facts, not verdicts — each flagged term links to its definition and the exact clause on this page. Absence of a badge means the term isn't currently flagged here, not that the page is risk-free.
Claude.ai – Developer Terms
Dataset: GenGA (Generative AI Governance Archive) — 11 AI services, 2025–present
GenGA (Generative AI Governance Archive) is this wiki’s Generative-AI dataset: raw policy snapshots captured by the Open Terms Archive
genai-euproject across 11 GenAI providers (ChatGPT, Claude.ai, DeepSeek, Google Generative AI Services, Le Chat, Llama API, Meta AI, Microsoft Copilot, Perplexity, Qwen Chat, xAI). Unlike PGAv2, GenGA has no pre-tagged risk scores — all risk scoring and keyword tagging on these pages is LLM-assigned by direct reading, spanning 2025–present.
Note: This page contains documented policy clauses. Risk assessment is qualitative and context-dependent. For analysis of patterns across platforms, see:
Source: sources/GenGA/Claude.ai/Developer Terms/ (raw Markdown captures, Open Terms Archive genai-eu dataset — no pre-tagged JSONL exists for this source)
Last updated: 2026-04-15
Note on methodology: No pre-computed
risk_score/keywordfields exist for this source. Keyword tags and clause analysis on this page are LLM-assigned, using the same rubric/vocabulary as the PGAv2 pages plus new GenAI-specific tags (royalty-free,improve services). Treat as first-pass analysis, not externally verified ground truth. This wiki’s risk-scoring system was retired project-wide on 2026-06-21 (seemethodology.md§4) — no page, GenGA included, computes or displays a numeric risk score.
Overview
1. What the document is: Anthropic’s “Software Directory Terms” and “Software Directory Policy” — governs third-party developers submitting MCP servers, Skills, plugins, and apps to Anthropic’s directories (Claude’s equivalent of OpenAI’s App Developer Terms / Connector ecosystem).
2. Input rights: Narrower than OpenAI’s equivalent. Developers grant Anthropic “non-exclusive, royalty-free, worldwide licenses to reproduce, display, and distribute any descriptions of the Software and Software documentation,” plus rights to display the developer’s branding “for the purpose of identifying and promoting” the Software. This only covers descriptions, metadata, and branding — not the Software’s actual runtime responses, unlike OpenAI’s Connector/App Response license, which does cover response content.
3. Output restrictions: §4 “Unsupported Use Cases” bars Software that generates images/video/audio as its main function (with a carve-out for design tools), serves ads, or handles financial transactions — these are restrictions on content type, not licensing. §1.F separately bars Software from querying “Claude’s memory, chat history, conversation summaries, or user-generated or uploaded files” — a data-minimization rule that protects users from third-party Software overreach, not a restriction on the developer.
4. Non-explicit predatory clauses: Anthropic can “collect and use functional metadata about your Software to improve the Anthropic Services, and to share such functional metadata with our users” — an “improve services”-style data-use right, though limited to metadata about how the Software runs, not its content or the underlying API responses.
5. Regulatory references: None found — no EU AI Act or GDPR mention. The document references Anthropic’s own Usage Policy, Supported Regions Policy, and Trademark Guidelines instead.
6. Regional variation: Not addressed in this document (it links to a policy listing “the countries and regions Anthropic currently supports,” but sets no separate regional terms here).
7. Key risk to users (developers, in this case): Low-to-moderate. The royalty-free license over descriptions/documentation/branding is narrower and lower-risk than OpenAI’s equivalent, since it doesn’t reach the Software’s actual runtime content. The bigger risk for developers is Anthropic’s right to change these terms at any time without notice, combined with an unconditional right to remove any Software “at any time for any reason… without liability to you.”
Flagged Keywords & Risks (LLM-assigned)
royalty-free— “non-exclusive, royalty-free, worldwide licenses to reproduce, display, and distribute any descriptions of the Software and Software documentation.” Why it matters: narrower than the equivalent clause in OpenAI’s Developer Terms — it only covers descriptions and branding, not runtime API responses — making this one of the lower-risk “royalty-free” clauses tracked in this dataset.improve services— “authorize Anthropic to… collect and use functional metadata about your Software to improve the Anthropic Services, and to share such functional metadata with our users.” Why it matters: a narrower, metadata-only version of the common “improve services” catch-all.
Regulatory & Research Context
Pandit et al. (2026) note that GenAI providers’ terms generally claim broad rights to use developer- and user-supplied material “for other purposes beyond” the immediate service; the metadata-scoped improve services clause here is a narrower instance of that pattern, applying to functional metadata about third-party Software rather than to the Software’s runtime content itself. Davidson et al. (2026), who include Anthropic among the providers they studied directly, would situate Anthropic’s unilateral, no-notice modification right and unconditional removal power (“at any time for any reason… without liability to you”) as the kind of unreviewable platform discretion their comparative analysis of regulatory gray areas highlights as a recurring feature of LLM Terms of Service.
Changes Summary
| Date | What changed |
|---|---|
| 2026-03-17 | Baseline version (first capture in this dataset). |
| 2026-04-15 | Cosmetic only: section-heading markdown style changed (dashes → equals signs); one MCP annotations-spec link updated from a “draft” version to a dated “2025-11-25” version. No wording change to any flagged clause. |
Version History
2026-03-17
- Explicit AI clause: YES
- Non-explicit predatory: YES
- Flagged keywords:
royalty-free,improve services
Clause: royalty-free, improve services
You further (i) agree to indemnify and hold Anthropic harmless from any claims, damages, or liabilities arising from or related to your Software or users’ interactions with it; (ii) grant Anthropic non-exclusive, royalty-free, worldwide licenses to reproduce, display, and distribute any descriptions of the Software and Software documentation provided by you or made available through the Software in connection with presenting the Software in the Directories and to display your name, trademarks, logos, and other branding materials associated with your Software in Directories for the purpose of identifying and promoting your Software, our Directories, and the Anthropic Services to our users and the public; and (iii) authorize Anthropic to review and test your Software in accordance with these terms, to collect and use functional metadata about your Software to improve the Anthropic Services, and to share such functional metadata with our users.
2026-04-15
royalty-free/improve services unchanged from 2026-03-17 — this date’s only change was heading-style formatting and a link-version update, neither affecting a flagged clause.