Skip to content
Browse logo

Browse MCP server

Which MCP servers are Browse's own, from the official registry and from asking its own domain, whether they answer, and what they offer. A program checked each of these, nobody copied them from a product page.

read 2026-10-26

← Back to Browse

Whose server it is
  • Registry nameorg.promptdna/promptdnaverified in the MCP registry

The registry name sits under promptdna.org. The registry only lets a publisher use that namespace after proving control of the domain through DNS or HTTP.

When we asked itchecked 2026-10-26
  • Stateanswerschecked 2026-10-26
  • Tools20
  • Server calls itselfpromptdna
  • Server version1.28.1
  • Protocol version2025-06-18

On 2026-10-26 the server answered initialize and tools/list without a sign-in. First answer in 205 ms.

Should an agent connect itNot proven either way yet
  • Whether its tools do the product's job has not been judged yet.
  • Checked on 1 of the 7 days it takes to judge reliability.

Once connected

  • Parameters are untyped or undescribed in 19 of 20 tools: check arguments before calling.

Said by rules from the measurements above. The likely first call is a model's pick from the tool list; read its description before relying on it.

Measured dailydaily checks: 1, the first on 2026-10-26
  • INITE MCP scoreafter a weeka score needs 7 daily checks: 1 so far
  • Answeredmeasuring1 of the 7 daily checks it takes to print a share
  • Median first answer194 ms
  • Space in an agent's context≈ 2,769 tokensthe definitions of all its tools (20), which an agent carries on every turn
  • Tool definitionsC · 68 of 100
  • Typed results0 of 20 toolsdeclared outputSchema: results come typed, read once
  • Parameters without a description: in 19 of 20
  • Not saying whether it only reads or can destroy: 20 of 20

What an agent can do with it

  • not sure · Generate D&D adventure modules on specific themes
  • not sure · Browse validated prompt blocks by category
  • not sure · Filter prompt blocks by approach or constraint

Each of the product's use cases, put to Jev against this server's tool list: can an agent do it with these tools?

Worth a look

None of the five warning signs in these definitions.

A program checks the server once a day: initialize and tools/list, no model. Tokens are the definitions' length at about four characters a token. The grade averages six checks of every tool's name, description, schema and annotations; it says how well the tools are described, not whether they work. The score weighs what was measured and nothing else: answering (25%), doing the product's job (25%), definitions (20%), context (10%), use (10%) and how readily a client connects (10%). A part that was not measured is left out, and a hidden instruction caps the score at 40. The state of vendors' MCP servers: method and numbers →

Tools it offers20 tools · read 2026-10-26
  • search_blocks

    Semantic/keyword search over blocks. Free for the first 30 calls/account/rolling hour, then 1 credit/call - max 50 results per call either way.

  • get_block

    Fetch a specific block by id. Costs 0.5 credits. Rate-limited to 40 calls/account (or per-IP if unauthenticated)/rolling hour.

  • compose_prompt

    Assemble a prompt from blocks for a task - the primary tool. Costs 5 credits.

  • get_collection

    Fetch a curated collection of blocks. Costs 0.5 credits.

  • find_compatible_blocks

    Given a block, return its requires/recommends/conflicts relationships. Costs 0.5 credits.

  • get_trending

    Top blocks by usage/rating, optionally scoped by domain/category. Costs 0.5 credits.

  • submit_block

    Submit a new block. Costs 2 credits, drawn from the submission-only bucket first.

  • submit_rating

    Rate a block 1-5. Always free.

  • fork_block

    Fork an existing block into a new lineage you own. Costs 2 credits.

  • get_block_versions

    Changelog for a block's lineage. Costs 0.5 credits.

  • get_benchmarks

    List published benchmark test cases, filterable by category and/or block_id, with aggregate pass rates. Costs 0.5 credits.

  • submit_benchmark

    Submit a new benchmark test case. Costs 2 credits, drawn from the submission-only bucket first. Always created pending - requires moderator approval before it's eligible to run or appears on get_benchmarks.

  • get_bounties

    List block bounties, filterable by category and/or status (defaults to open). Costs 0.5 credits.

  • report_injection_attempt

    Report a suspected prompt-injection block. Always free.

  • submit_dmca_takedown

    File a DMCA takedown notice. Always free, no account required.

  • submit_dmca_counter_notice

    File a counter-notice against a DMCA takedown. Always free.

  • get_my_appealable_penalties

    List your own credit penalties still eligible for appeal. Always free.

  • submit_appeal

    Appeal one of your own credit penalties. Costs a 10-credit bond (purchased/earned credits only), refunded if approved.

  • register_agent

    Step 1 of autonomous self-registration - no API key required. Costs $0.01 USDC on Base (anti-spam fee, not a wallet-ownership proof). Returns a nonce; sign it and call verify_registration within 5 minutes to get an API key.

  • verify_registration

    Step 2 of autonomous self-registration - no API key required. Proves control of the wallet used in register_agent via an ECDSA signature over its nonce, then issues an API key. Always free.

How to connect

Add the server to your MCP client config. Claude Desktop, Cursor and most other clients read this shape:

{
  "mcpServers": {
    "promptdna": {
      "type": "http",
      "url": "https://mcp.promptdna.org/v1/"
    }
  }
}

Add it

Claude Code, in a terminal

claude mcp add --transport http promptdna https://mcp.promptdna.org/v1/

Add to CursorAdd to VS Code

claude.ai, ChatGPT and other clients: add https://mcp.promptdna.org/v1/ as a custom connector in the settings.

OpenAPI spec

106 paths, read 2026-10-26. Open the spec ↗

Each line is something a program found when it went looking: the registry entry, the server's own answer, a spec that parses. A missing line means we did not find it, not that it does not exist.