Prep Room MCP server

• Addresshttps://preproom.ai/api/mcpnot in the official registry: answered when asked

No entry in the official MCP registry names this server. It answered on preproom.ai when the Atlas asked at the address servers are usually mounted on, so it is the vendor's own. Clients and catalogues that list servers from the registry do not show it.

• Stateanswerschecked 2026-10-27
• Tools14
• Server calls itselfpreproom
• Server version1.0.0
• Protocol version2025-06-18

On 2026-10-27 the server answered initialize and tools/list without a sign-in. First answer in 1409 ms.

• Its tools do none of the product's use cases we checked: write and run coding solutions with hidden tests; check a resume against a specific job description.
• Checked on 2 of the 7 days it takes to judge reliability.

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.

• INITE MCP scoreafter a weeka score needs 7 daily checks: 2 so far
• What an agent can do with it0 of 2 use cases
• Answeredmeasuring2 of the 7 daily checks it takes to print a share
• Median first answer1345 ms
• Space in an agent's context≈ 2,411 tokensthe definitions of all its tools (14), which an agent carries on every turn
• Tool definitionsA · 94 of 100
• Typed results0 of 14 toolsdeclared outputSchema: results come typed, read once

• Parameters without a description: in 5 of 14

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 →

• find_questions

  Finds real technical interview questions from Prep Room: AI-assisted engineering (reviewing AI-written code, agents, prompts and specs, working with an assistant), system design, and coding (algorithms, code review, on-call incidents…

• todays_question

  Today's Prep Room question: the same one for everyone today, a good warm-up. Returns the full brief, ready to ask.

• get_question

  The full brief for one question: the prompt, its format, level and usual time, and what the answer looks like. A coding question carries its signature, starter code in the language asked for, and examples; a multi-file question carries…

• get_hint

  One hint for a question, from a ladder that goes from a nudge to an outline. Ask for rung 1 first and only go further when the person is still stuck.

• get_answer

  A question's model answer in words, with the checklist a strong answer covers, the approaches from weakest to strongest, the steps of the best one, follow-up questions, common pitfalls and what each level is expected to show. Use it to…

• find_guides

  Finds short concept guides on the ideas behind interview questions (consistent hashing, idempotency, reviewing AI-written code, dynamic programming and many more). Search by words, area or topic, or pass a question id for the guides that…

• read_guide

  One section of a concept guide as text, plus questions to practice it. Teach it in your own words with an example; ask for the next section only when the person wants more.

• find_paths

  Prep Room's study paths: ordered sets of questions from warm-up to under-pressure, such as "System design foundations" or "Reviewing AI-written code". With no path, lists them; with one, gives its stages and questions.

• my_progress

  The person's Prep Room progress, the same as on the site: each track and the topics they have started as eight squares (■ solved, ◩ attempted, □ open), mastery levels, and the next question on each path they started. Needs a linked account.

• whats_due

  What the person should do next: solved questions due for review today and later this week, and the next days of their study plan if they set one. Needs a linked account.

• mark_question

  Records that the person worked on a question, so it shows on their Prep Room progress. "attempted" fits any question. "solved" fits questions answered in words (design, code review, on-call, judgment calls) once they have covered the…

• record_review

  For a solved question that came back for review: did the person get it again? A pass moves the next review later; a miss brings it back tomorrow. Needs a linked account with practice allowed.

• set_study_plan

  Lays a study path out on the days left before an interview, two to five days a week, the same plan the site shows. Replaces any plan the person had. Needs a linked account with practice allowed.

• get_reference

  The reference solution in code for a coding or SQL question, for Prep Room plan members. For everyone else the approach in words (get_answer) is open, and the code is on the question's page with the plan.