Skip to content
Agentmart
Submit a tool

Draft, pending legal review. This document has not been reviewed by a lawyer yet and may change before it takes effect. Questions: tell us.

Acceptable Use Policy

Version 0.1 · last updated 2026-10-26

Draft for legal review. Not in force. Clauses marked [LAWYER: ...] or [RAHIL: ...] are open.

1. Scope

This policy is part of the Terms of Service (/legal/terms). It applies to everyone who uses Agent Mart:

  • visitors
  • account holders
  • publishers
  • API and MCP users
  • CLI users

It covers anything you submit, including listings, reviews, replies, stacks, reports and profile details.

2. No malware or harmful tools

Do not submit, list or link to software that:

  • is malware, ransomware, spyware or a backdoor
  • steals credentials, keys or data
  • hides what it does, or does things its description and permissions do not disclose
  • injects instructions into AI agents to make them act against their users (prompt injection or tool poisoning)
  • mines cryptocurrency without clear consent

Security research tools are welcome if their purpose and risks are described plainly. [LAWYER: line between offensive security tools and malware for listing purposes.]

3. No impersonation or false claims

Do not:

  • pretend to be another person, company or project
  • claim a listing you do not maintain
  • submit a copy of someone else's tool as your own
  • use names or logos that mislead people about who publishes a tool
  • suggest that Agent Mart audited, certified or endorses something it did not

4. No fake reviews or manipulation

Do not:

  • write reviews of tools you did not use, or of your own tools or your employer's
  • write reviews of competitors to harm them
  • pay, reward or pressure anyone to write, change or remove a review
  • use several accounts, scripts or coordinated groups to post reviews, stacks or reports

5. No attempts to manipulate rank or grades

Do not try to change how a listing ranks or what grade it gets by any means other than improving the tool itself. This includes:

  • inflating stars, downloads, installs or telemetry
  • creating fake listings or duplicates
  • hiding code from our scanners
  • serving different content to our crawlers than to users

Money never changes organic rank or grades. No one can pay us for a better rank or grade, so do not offer.

6. No scraping beyond the API terms

Use the public API or MCP server to get catalog data, within their published limits and terms (/developers#terms). Do not:

  • crawl the website at a rate that harms it
  • get around rate limits, for example by rotating IP addresses or keys
  • copy or republish the catalog, or substantial parts of it, without a license

Search engines and AI crawlers that follow our robots.txt are welcome. [LAWYER: confirm the robots.txt and API-terms interplay.]

7. API and MCP fair use

  • Keep your API keys secret; one key per app or person.
  • Stay within the limits for your tier. Keys that hit their limit on three days in a row are paused automatically. Ask us to restore one.
  • Cache responses where you can, and show "Data from Agent Mart" with a link where you display our data.
  • Do not use the API to harm listed projects or their maintainers.

8. No abuse of people or the service

Do not:

  • harass, threaten or dox anyone, including maintainers and reviewers
  • post hateful, sexual or violent content
  • post spam, unsolicited ads or link schemes
  • put personal data of others in reviews, stacks or replies without their permission
  • probe, scan or attack Agent Mart outside the Security Policy (/legal/security-policy)
  • bypass bot checks (Cloudflare Turnstile), form limits or bans
  • use the service to break any law, including export control and sanctions laws [LAWYER: wording]

9. Reporting

Report anything that breaks this policy:

  • on any listing, with "Report a problem"
  • with the report form at https://agentmart.dev/submit?kind=report
  • for vulnerabilities, see the Security Policy

We aim to answer security reports within 24 hours.

10. What we may do

Depending on how serious it is, we may:

  • remove or hide content (we mark it removed rather than deleting it)
  • hold reviews for manual moderation
  • revoke claims
  • pause or revoke API keys
  • suspend accounts
  • report illegal content to the authorities

We will tell you why, where we can, and you can ask us to review the decision by replying to our message or by using the report form. [LAWYER: statement of reasons and appeal requirements (EU Digital Services Act if it applies; Indian IT Rules 2021 if we are an intermediary).]

Open questions for the lawyer

  1. [LAWYER] Where to draw the line between offensive security tools and malware (section 2).
  2. [LAWYER] Interplay of robots.txt, the API terms and the anti-scraping clause (section 6).
  3. [LAWYER] Export control and sanctions wording (section 8).
  4. [LAWYER] Statement of reasons and appeal duties under the EU Digital Services Act and the Indian IT Rules 2021 (section 10).