Finished jobs, paid per result.

Each block does one job: input in, result out. Run it from this page, from your own software, or from an AI agent. Every block shows its price per result before you run it.

See the blocks Ways to run them

THE STORE

Blocks you can run today.

To try a block, paste a run key. The operator's run key is in crew /root/dev/blocks/.env, BLOCKS_RUN_KEYS.

appsim 0.2.0

Outside sites it contacts: api.anthropic.com

View manifest

read_job_ad — Must-haves, nice-to-haves and deal-breakers from one job ad, each with its verbatim quote. up to $0.05 per unit Inputs: Job ad Try
read_resume — Titles with dates and claims from one CV or service description, each with its verbatim quote. up to $0.05 per unit Inputs: CV or service description Try
match_report — Each job-ad requirement checked against the CV and message's exact wording; counts met, recruiter notice, and fixes. up to $0.05 per unit Inputs: Job ad, CV or service description, Cover message Try

From your software

Replace <YOUR_RUN_KEY> with your run key. The operator's run key is in crew /root/dev/blocks/.env, BLOCKS_RUN_KEYS; make more on the run keys page.

Address: https://brickbank.cc/b/appsim

Call it:

# Use a new code for each new run. Send the same code again when you retry, and you get the first run back, never charged twice.
curl -X POST https://brickbank.cc/b/appsim/v1/run/read_job_ad -H "Authorization: Bearer <YOUR_RUN_KEY>" -H "Idempotency-Key: <NEW_CODE_PER_RUN>" -d '{"items": [{"id": "example", "input": {"job_ad": "Senior Backend Engineer (Python) \u2014 Northwind Analytics, remote (EU)\n\nWe build the data platform behind retail demand forecasting.\n\nMust have:\n- 5+ years of professional Python experience\n- Production experience with PostgreSQL\n- Designing and running REST APIs at scale\n\nNice to have:\n- Kubernetes\n- Experience with event streaming (Kafka)\n\nWe cannot sponsor visas; you must be able to work in the EU."}}]}'

MCP address: https://brickbank.cc/b/appsim/mcp (tools)

Claude Code:

claude mcp add --transport http appsim https://brickbank.cc/b/appsim/mcp --header "Authorization: Bearer <YOUR_RUN_KEY>"

.mcp.json (other MCP clients):

{
  "mcpServers": {
    "appsim": {
      "type": "http",
      "url": "https://brickbank.cc/b/appsim/mcp",
      "headers": {
        "Authorization": "Bearer <YOUR_RUN_KEY>"
      }
    }
  }
}

Run keys page (needs a run key)

read_job_ad

Must-haves, nice-to-haves and deal-breakers from one job ad, each with its verbatim quote.

Job ad *

Example call

# Use a new code for each new run. Send the same code again when you retry, and you get the first run back, never charged twice.
curl -X POST https://brickbank.cc/b/appsim/v1/run/read_job_ad -H "Authorization: Bearer <YOUR_RUN_KEY>" -H "Idempotency-Key: <NEW_CODE_PER_RUN>" -d '{"items": [{"id": "example", "input": {"job_ad": "Senior Backend Engineer (Python) \u2014 Northwind Analytics, remote (EU)\n\nWe build the data platform behind retail demand forecasting.\n\nMust have:\n- 5+ years of professional Python experience\n- Production experience with PostgreSQL\n- Designing and running REST APIs at scale\n\nNice to have:\n- Kubernetes\n- Experience with event streaming (Kafka)\n\nWe cannot sponsor visas; you must be able to work in the EU."}}]}'

read_resume

Titles with dates and claims from one CV or service description, each with its verbatim quote.

CV or service description *

Example call

# Use a new code for each new run. Send the same code again when you retry, and you get the first run back, never charged twice.
curl -X POST https://brickbank.cc/b/appsim/v1/run/read_resume -H "Authorization: Bearer <YOUR_RUN_KEY>" -H "Idempotency-Key: <NEW_CODE_PER_RUN>" -d '{"items": [{"id": "example", "input": {"resume": "Backend Engineer, Contoso Retail \u2014 2019 to present\n- 6 years building Python services (FastAPI, SQLAlchemy)\n- Designed the public REST API serving 40M requests a day\n- Ran PostgreSQL clusters in production, including zero-downtime migrations\n\nSoftware Developer, Fabrikam Logistics \u2014 2016 to 2019\n- Built shipment tracking pipelines on Kafka\n\nBased in Lisbon, Portugal; EU citizen."}}]}'

match_report

Each job-ad requirement checked against the CV and message's exact wording; counts met, recruiter notice, and fixes.

Job ad *

CV or service description *

Cover message *

Example call

# Use a new code for each new run. Send the same code again when you retry, and you get the first run back, never charged twice.
curl -X POST https://brickbank.cc/b/appsim/v1/run/match_report -H "Authorization: Bearer <YOUR_RUN_KEY>" -H "Idempotency-Key: <NEW_CODE_PER_RUN>" -d '{"items": [{"id": "example", "input": {"job_ad": "Senior Backend Engineer (Python) \u2014 Northwind Analytics, remote (EU)\n\nWe build the data platform behind retail demand forecasting.\n\nMust have:\n- 5+ years of professional Python experience\n- Production experience with PostgreSQL\n- Designing and running REST APIs at scale\n\nNice to have:\n- Kubernetes\n- Experience with event streaming (Kafka)\n\nWe cannot sponsor visas; you must be able to work in the EU.", "resume": "Backend Engineer, Contoso Retail \u2014 2019 to present\n- 6 years building Python services (FastAPI, SQLAlchemy)\n- Designed the public REST API serving 40M requests a day\n- Ran PostgreSQL clusters in production, including zero-downtime migrations\n\nSoftware Developer, Fabrikam Logistics \u2014 2016 to 2019\n- Built shipment tracking pipelines on Kafka\n\nBased in Lisbon, Portugal; EU citizen.", "message": "Hello Northwind team,\n\nI'd like to apply for the Senior Backend Engineer role. For the past six years I have built Python\nservices and the REST API at Contoso Retail, and I run its PostgreSQL clusters in production.\n\nBest regards"}}]}'

bossroster 0.1.0

Outside sites it contacts: api.anthropic.com, app.icypeas.com

View manifest

From your software

Replace <YOUR_RUN_KEY> with your run key. The operator's run key is in crew /root/dev/blocks/.env, BLOCKS_RUN_KEYS; make more on the run keys page.

Address: https://brickbank.cc/b/bossroster

Call it:

# Use a new code for each new run. Send the same code again when you retry, and you get the first run back, never charged twice.
curl -X POST https://brickbank.cc/b/bossroster/v1/run/resolve -H "Authorization: Bearer <YOUR_RUN_KEY>" -H "Idempotency-Key: <NEW_CODE_PER_RUN>" -d '{"items": [{"id": "example", "input": {"domain": "stripe.com"}}]}'

MCP address: https://brickbank.cc/b/bossroster/mcp (tools)

Claude Code:

claude mcp add --transport http bossroster https://brickbank.cc/b/bossroster/mcp --header "Authorization: Bearer <YOUR_RUN_KEY>"

.mcp.json (other MCP clients):

{
  "mcpServers": {
    "bossroster": {
      "type": "http",
      "url": "https://brickbank.cc/b/bossroster/mcp",
      "headers": {
        "Authorization": "Bearer <YOUR_RUN_KEY>"
      }
    }
  }
}

Run keys page (needs a run key)

ask

Answers a question about a resolved company with evidence and honest abstention. count_function: how many people work in a function (talent, engineering), worldwide or in one country, from free counts only; 1 or 2 read as a range, never a person. has_head: whether the function has a head, widening country, region, world, with a reason per step; the CEO counts only under 50 staff; an absence is stated only above the calibrated floor, else unknown; never a person. dm_for_posting: the person who leads the function in a country, named only for a declared purpose and only when the ranker and the title dictionary agree, at $0.10, refused behind the privacy gates; nobody acceptable is not_found and free. org_picture: the company picture at $1, a free skyline of every function by leaders and staff, a seat verdict per function and, for up to 5 functions the caller names with a declared purpose, their head; never a reporting line. max_credits 0 quotes and spends nothing; a fresh re-ask is free.

Canonical company key (from resolve) *

lpcritic 0.1.0

Outside sites it contacts: api.anthropic.com

View manifest

line_check — Checks one headline, subhead or CTA line: its promise rung 0-4 (concrete = 3+), what it says back, whether a competitor could say it, and rule-cited findings. No rewrites. $0.02 per unit Inputs: Headline or line, Role Try
hero_check — Critiques a landing page's first screen (URL or md/html upload) at 1280 and 390 px: a stranger's cold read, the hero's promises and rule-cited findings, top 3 first. Never judges whether the product delivers. $0.06 per unit Inputs: Landing page URL, Uploaded draft Try
page_critique — Critiques a whole landing page (URL or md/html upload): its cold read and every promise with its proof and whether the page keeps it. Never judges whether the product delivers. $0.31 per unit Inputs: Landing page URL, Uploaded draft Try

From your software

Replace <YOUR_RUN_KEY> with your run key. The operator's run key is in crew /root/dev/blocks/.env, BLOCKS_RUN_KEYS; make more on the run keys page.

Address: https://brickbank.cc/b/lpcritic

Call it:

# Use a new code for each new run. Send the same code again when you retry, and you get the first run back, never charged twice.
curl -X POST https://brickbank.cc/b/lpcritic/v1/run/line_check -H "Authorization: Bearer <YOUR_RUN_KEY>" -H "Idempotency-Key: <NEW_CODE_PER_RUN>" -d '{"items": [{"id": "example", "input": {"line": "Automate with confidence", "role": "headline"}}]}'

MCP address: https://brickbank.cc/b/lpcritic/mcp (tools)

Claude Code:

claude mcp add --transport http lpcritic https://brickbank.cc/b/lpcritic/mcp --header "Authorization: Bearer <YOUR_RUN_KEY>"

.mcp.json (other MCP clients):

{
  "mcpServers": {
    "lpcritic": {
      "type": "http",
      "url": "https://brickbank.cc/b/lpcritic/mcp",
      "headers": {
        "Authorization": "Bearer <YOUR_RUN_KEY>"
      }
    }
  }
}

Run keys page (needs a run key)

instantly 1.0.0

Outside sites it contacts: api.instantly.ai

View manifest

campaign.enrol — Enrol a lead into an Instantly campaign (and optionally a list) by email $0.00 per unit Inputs: campaign_id, contact_id, email, list_id Try
lead.add — Add a lead to an Instantly campaign or list with its names, company fields and custom variables $0.00 per unit Inputs: email, campaign_id, list_id, first_name, last_name, website, company_name, custom_variables Try
lead.list — List the leads of an Instantly campaign with their delivery status, paging Instantly internally $0.00 per unit Inputs: campaign_id, limit Try
campaign.sending_status — Read an Instantly campaign's daily send limit, how much of it is used, and whether it is hit $0.00 per unit Inputs: campaign_id Try

From your software

Replace <YOUR_RUN_KEY> with your run key. The operator's run key is in crew /root/dev/blocks/.env, BLOCKS_RUN_KEYS; make more on the run keys page.

Address: https://brickbank.cc/b/instantly

campaign.enrol has no example input yet, so there is no ready-made call: try it to see its inputs.

MCP address: https://brickbank.cc/b/instantly/mcp (tools)

Claude Code:

claude mcp add --transport http instantly https://brickbank.cc/b/instantly/mcp --header "Authorization: Bearer <YOUR_RUN_KEY>"

.mcp.json (other MCP clients):

{
  "mcpServers": {
    "instantly": {
      "type": "http",
      "url": "https://brickbank.cc/b/instantly/mcp",
      "headers": {
        "Authorization": "Bearer <YOUR_RUN_KEY>"
      }
    }
  }
}

Run keys page (needs a run key)

lemlist 1.0.0

Outside sites it contacts: api.lemlist.com

View manifest

campaign.enrol — Enrol a lead into a lemlist campaign by email $0.00 per unit Inputs: campaign_id, contact_id, email, list_id Try
campaigns.list — List the account's lemlist campaigns $0.00 per unit Inputs: none Try
campaigns.inventory — List the account's lemlist campaigns with status and the vendor's boolean flags $0.00 per unit Inputs: none Try
campaign.preflight — Whether a lemlist campaign can dispatch a lead right now, without adding one $0.00 per unit Inputs: campaign_id Try
lead.add — Add a lead (email and/or LinkedIn URL, names, custom variables) to a lemlist campaign that can dispatch it $0.00 per unit Inputs: campaign_id, email, linkedin_url, first_name, last_name, variables Try
activities.pull — The lemlist activity feed (sends, opens, replies, ...) since an instant, as raw rows $0.00 per unit Inputs: since Try
seats.count — How many distinct lemlist users (seats) the account has $0.00 per unit Inputs: none Try
key.validate — Whether a lemlist API key is valid, free of charge $0.00 per unit Inputs: none Try
employment.confirm — Whether lemlist's people database places a LinkedIn profile at the given company $0.00 per unit Inputs: full_name, first_name, last_name, company_domain, company_name, company_urn, linkedin_url Try
people.stated_employment — The current employer lemlist's people database states for a LinkedIn profile $0.00 per unit Inputs: linkedin_url Try

From your software

Replace <YOUR_RUN_KEY> with your run key. The operator's run key is in crew /root/dev/blocks/.env, BLOCKS_RUN_KEYS; make more on the run keys page.

Address: https://brickbank.cc/b/lemlist

campaign.enrol has no example input yet, so there is no ready-made call: try it to see its inputs.

MCP address: https://brickbank.cc/b/lemlist/mcp (tools)

Claude Code:

claude mcp add --transport http lemlist https://brickbank.cc/b/lemlist/mcp --header "Authorization: Bearer <YOUR_RUN_KEY>"

.mcp.json (other MCP clients):

{
  "mcpServers": {
    "lemlist": {
      "type": "http",
      "url": "https://brickbank.cc/b/lemlist/mcp",
      "headers": {
        "Authorization": "Bearer <YOUR_RUN_KEY>"
      }
    }
  }
}

Run keys page (needs a run key)

whatblock 0.1.0

Outside sites it contacts: api.anthropic.com

View manifest

company.what_it_does — What a company actually does — what it sells, to whom, where — in 1-3 plain third-person sentences, from its own LinkedIn About and website texts. No self-praise, mission or founding story; never invents facts. $0.00060 per unit Inputs: company_name, profile_text, website_text, industry Try

From your software

Replace <YOUR_RUN_KEY> with your run key. The operator's run key is in crew /root/dev/blocks/.env, BLOCKS_RUN_KEYS; make more on the run keys page.

Address: https://brickbank.cc/b/whatblock

Call it:

# Use a new code for each new run. Send the same code again when you retry, and you get the first run back, never charged twice.
curl -X POST https://brickbank.cc/b/whatblock/v1/run/company.what_it_does -H "Authorization: Bearer <YOUR_RUN_KEY>" -H "Idempotency-Key: <NEW_CODE_PER_RUN>" -d '{"items": [{"id": "example", "input": {"company_name": "Sandline HR", "profile_text": "We are a family of passionate, result-driven people, proudly serving since 2015! Our journey began with a dream to change how businesses grow in the region.", "website_text": "Payroll outsourcing | Work visas & PRO services | Recruitment | For startups and SMEs across Dubai and Abu Dhabi | 300+ clients", "industry": "Human Resources Services"}}]}'

MCP address: https://brickbank.cc/b/whatblock/mcp (tools)

Claude Code:

claude mcp add --transport http whatblock https://brickbank.cc/b/whatblock/mcp --header "Authorization: Bearer <YOUR_RUN_KEY>"

.mcp.json (other MCP clients):

{
  "mcpServers": {
    "whatblock": {
      "type": "http",
      "url": "https://brickbank.cc/b/whatblock/mcp",
      "headers": {
        "Authorization": "Bearer <YOUR_RUN_KEY>"
      }
    }
  }
}

Run keys page (needs a run key)

company.what_it_does

What a company actually does — what it sells, to whom, where — in 1-3 plain third-person sentences, from its own LinkedIn About and website texts. No self-praise, mission or founding story; never invents facts.

company_name *

profile_text

website_text

industry

Example call

# Use a new code for each new run. Send the same code again when you retry, and you get the first run back, never charged twice.
curl -X POST https://brickbank.cc/b/whatblock/v1/run/company.what_it_does -H "Authorization: Bearer <YOUR_RUN_KEY>" -H "Idempotency-Key: <NEW_CODE_PER_RUN>" -d '{"items": [{"id": "example", "input": {"company_name": "Sandline HR", "profile_text": "We are a family of passionate, result-driven people, proudly serving since 2015! Our journey began with a dream to change how businesses grow in the region.", "website_text": "Payroll outsourcing | Work visas & PRO services | Recruitment | For startups and SMEs across Dubai and Abu Dhabi | 300+ clients", "industry": "Human Resources Services"}}]}'

datagate-people 0.1.0

Outside sites it contacts: api.hunter.io, api.leadmagic.io, app.bettercontact.rocks, app.icypeas.com

View manifest

From your software

Replace <YOUR_RUN_KEY> with your run key. The operator's run key is in crew /root/dev/blocks/.env, BLOCKS_RUN_KEYS; make more on the run keys page.

Address: https://brickbank.cc/b/datagate-people

bettercontact.email.find has no example input yet, so there is no ready-made call: try it to see its inputs.

MCP address: https://brickbank.cc/b/datagate-people/mcp (tools)

Claude Code:

claude mcp add --transport http datagate-people https://brickbank.cc/b/datagate-people/mcp --header "Authorization: Bearer <YOUR_RUN_KEY>"

.mcp.json (other MCP clients):

{
  "mcpServers": {
    "datagate-people": {
      "type": "http",
      "url": "https://brickbank.cc/b/datagate-people/mcp",
      "headers": {
        "Authorization": "Bearer <YOUR_RUN_KEY>"
      }
    }
  }
}

Run keys page (needs a run key)

icypeas.people.rank

Pick the right decision-maker from people the caller already holds, for one posting: no vendor call, free. Drops a person from another function or country (a regional title that covers the posting's country passes), takes a CEO, MD or founder only when the company has no head of the posting's function, and returns the rest best first, each with a `pick_role` (posting_manager, function_head, small_company_ceo, executive or senior_contact), a `pick_reason` code and a `pick_why` sentence. A posting's `exclude_titles` drop a person whose title holds one of those words, and its `seniority` bands rank a person further from them lower (`off_seniority_band`). Nobody acceptable is a `not_found` miss that lists every rejected person's title, country and reason, or with `not_found_as` data an ok answer carrying that as fields. `strict` false lets generalists through.

What the pick is for: the job posting, or fixed client titles for a company-only pipeline

candidates *

Example call

# Use a new code for each new run. Send the same code again when you retry, and you get the first run back, never charged twice.
curl -X POST https://brickbank.cc/b/datagate-people/v1/run/icypeas.people.rank -H "Authorization: Bearer <YOUR_RUN_KEY>" -H "Idempotency-Key: <NEW_CODE_PER_RUN>" -d '{"items": [{"id": "example", "input": {"posting": {"posting_country": "United Arab Emirates", "posting_function": "product"}, "candidates": [{"full_name": "Jane Doe", "title": "Head of Product, GCC"}]}}]}'

icypeas.people.find_ranked

People at one company (domain, name or LinkedIn company URN) or anywhere, for the first of `title_any` (1-10 titles, most likely first) that finds someone: one Icypeas People Search per title, 10 people by default and up to 200, stopping at the first title with people left after the locations filter. With `tiers` (each a ranked title list) the same first-hit search runs once per tier and then once for `title_any` as the last tier, `limit` applying per tier; a person already found by an earlier tier is dropped by LinkedIn URL, and each person carries its 0-based `tier`. A title is not started once another search would not fit inside the time budget: the tiers finished by then are returned, or a `not_found` miss when none found anyone. `title_none` words and the posting's `exclude_titles` are sent to the vendor as job-title excludes, so rows with those words are never read or billed; locations stay a filter on the returned page. Every row read is billed, including those the locations filter drops. Each person carries the vendor's `total` match count for the title that matched. With a `posting` the shared decision-maker picker also reads the people: `picker` shadow (the default) logs its pick beside this answer and changes nothing; `on` returns only the people it accepts, each with a `pick_role`, `pick_reason` and `pick_why`, moves to the next title when a page holds nobody acceptable, and answers a `not_found` that is not billed when every person read was rejected. A CEO, MD or founder is taken only when the company has no head of the posting's function: a free count the gateway keeps 30 days for every consumer answers that, the posting's `function_head_exists` overrides it and its `employee_count` decides when neither is known.

company_domain

company_name

company_urn

Job titles, most likely first *

What the pick is for: the job posting, or fixed client titles for a company-only pipeline

Example call

# Use a new code for each new run. Send the same code again when you retry, and you get the first run back, never charged twice.
curl -X POST https://brickbank.cc/b/datagate-people/v1/run/icypeas.people.find_ranked -H "Authorization: Bearer <YOUR_RUN_KEY>" -H "Idempotency-Key: <NEW_CODE_PER_RUN>" -d '{"items": [{"id": "example", "input": {"company_domain": "acme.com", "title_any": ["CEO", "Founder"]}}]}'

datagate-scrape 0.1.0

Outside sites it contacts: api.apify.com, api.ashbyhq.com, api.brightdata.com, api.eu.lever.co, api.harvestapi.io, api.lever.co, boards-api.greenhouse.io, www.reddit.com

View manifest

From your software

Replace <YOUR_RUN_KEY> with your run key. The operator's run key is in crew /root/dev/blocks/.env, BLOCKS_RUN_KEYS; make more on the run keys page.

Address: https://brickbank.cc/b/datagate-scrape

apify.actor_run has no example input yet, so there is no ready-made call: try it to see its inputs.

MCP address: https://brickbank.cc/b/datagate-scrape/mcp (tools)

Claude Code:

claude mcp add --transport http datagate-scrape https://brickbank.cc/b/datagate-scrape/mcp --header "Authorization: Bearer <YOUR_RUN_KEY>"

.mcp.json (other MCP clients):

{
  "mcpServers": {
    "datagate-scrape": {
      "type": "http",
      "url": "https://brickbank.cc/b/datagate-scrape/mcp",
      "headers": {
        "Authorization": "Bearer <YOUR_RUN_KEY>"
      }
    }
  }
}

Run keys page (needs a run key)

datagate-crm 0.1.0

Outside sites it contacts: api.clay.com, api.hubapi.com, api.instantly.ai

View manifest

From your software

Replace <YOUR_RUN_KEY> with your run key. The operator's run key is in crew /root/dev/blocks/.env, BLOCKS_RUN_KEYS; make more on the run keys page.

Address: https://brickbank.cc/b/datagate-crm

bullhorn.key.validate has no example input yet, so there is no ready-made call: try it to see its inputs.

MCP address: https://brickbank.cc/b/datagate-crm/mcp (tools)

Claude Code:

claude mcp add --transport http datagate-crm https://brickbank.cc/b/datagate-crm/mcp --header "Authorization: Bearer <YOUR_RUN_KEY>"

.mcp.json (other MCP clients):

{
  "mcpServers": {
    "datagate-crm": {
      "type": "http",
      "url": "https://brickbank.cc/b/datagate-crm/mcp",
      "headers": {
        "Authorization": "Bearer <YOUR_RUN_KEY>"
      }
    }
  }
}

Run keys page (needs a run key)

datagate-notify 0.1.0

Outside sites it contacts: api.mailgun.net, api.telegram.org, hooks.slack.com

View manifest

From your software

Replace <YOUR_RUN_KEY> with your run key. The operator's run key is in crew /root/dev/blocks/.env, BLOCKS_RUN_KEYS; make more on the run keys page.

Address: https://brickbank.cc/b/datagate-notify

mailgun.notify.send has no example input yet, so there is no ready-made call: try it to see its inputs.

MCP address: https://brickbank.cc/b/datagate-notify/mcp (tools)

Claude Code:

claude mcp add --transport http datagate-notify https://brickbank.cc/b/datagate-notify/mcp --header "Authorization: Bearer <YOUR_RUN_KEY>"

.mcp.json (other MCP clients):

{
  "mcpServers": {
    "datagate-notify": {
      "type": "http",
      "url": "https://brickbank.cc/b/datagate-notify/mcp",
      "headers": {
        "Authorization": "Bearer <YOUR_RUN_KEY>"
      }
    }
  }
}

Run keys page (needs a run key)

datagate-search 0.1.0

Outside sites it contacts: api.anthropic.com, api.company-information.service.gov.uk, api.exa.ai

View manifest

From your software

Replace <YOUR_RUN_KEY> with your run key. The operator's run key is in crew /root/dev/blocks/.env, BLOCKS_RUN_KEYS; make more on the run keys page.

Address: https://brickbank.cc/b/datagate-search

anthropic.llm.field has no example input yet, so there is no ready-made call: try it to see its inputs.

MCP address: https://brickbank.cc/b/datagate-search/mcp (tools)

Claude Code:

claude mcp add --transport http datagate-search https://brickbank.cc/b/datagate-search/mcp --header "Authorization: Bearer <YOUR_RUN_KEY>"

.mcp.json (other MCP clients):

{
  "mcpServers": {
    "datagate-search": {
      "type": "http",
      "url": "https://brickbank.cc/b/datagate-search/mcp",
      "headers": {
        "Authorization": "Bearer <YOUR_RUN_KEY>"
      }
    }
  }
}

Run keys page (needs a run key)

dibs 0.1.0

Outside sites it contacts: api.anthropic.com, api.name.com, api.porkbun.com, cloudflare-dns.com, crates.io, data.iana.org, dns.google, domainfinder.nic.at, github.com, pubapi.registry.google, pypi.org, rdap.identitydigital.services, rdap.nic.so, rdap.norid.no, rdap.registry.co, rdap.verisign.com, registry.npmjs.org, tld-rdap.verisign.com, whois.nic.at, whois.nic.io, whois.nic.sh, whois.nic.so

View manifest

names.suggest — Product-name candidates for a bonsai vision: name, why (citing V-ids), length, pronounceable, notes. $0.06 per unit Inputs: Vision, Current Name, Count, Style Try
naming.steer — The next round for a product card, a client-held envelope and this product's marks (`[{name, label, $1.47 per unit Inputs: Product, State, Marks Try
naming.critic — For each name the founder liked, in the order given: the strongest objection, its severity and a $0.25 per unit Inputs: Product, Names Try
naming.fit — For each name, in the order given: its fit to the card's vision, the passage it speaks to, why, a gloss. $0.36 per unit Inputs: Product, Names Try
naming.internal — Names for an internal tool, best first, each with how it was made and, when generated, the blind guess $0.08 per unit Inputs: Does, Siblings, Taken, Wraps, Family, Count Try
domains.check — Which `<name>.<tld>` domains are free: every name on `ALWAYS_TLDS` (.com, .no, .cc, .co, .at) first, $0.00 per unit Inputs: Names, Tlds Try
domains.ladder — Each name's best domain-ladder rung (free .com, free domain hack, free other TLD), with the $0.40 per unit Inputs: Names, Product, Tlds Try
naming.clear — Clearance for up to 8 finalists of a product card: handles (GitHub, npm, PyPI, crates; X and Instagram by $0.05 per unit Inputs: Product, Names Try
naming.dossier — The decision page for the names the founder starred or liked: the Pareto front of mark × domain path, a $0.00 per unit Inputs: Finalists, Ladder, Clearance, Classes, Prices, Critique, Funnel Try
domains.watch — Each domain's RDAP statuses, expiry, days left and drop phase now, asked of its registry alone. $0.00 per unit Inputs: Domains Try
domains.price — A registrar's price, premium flag and renewal for finalist domains (the caller's Porkbun keys), beside the registry verdict. $0.00 per unit Inputs: Domains Try

From your software

Replace <YOUR_RUN_KEY> with your run key. The operator's run key is in crew /root/dev/blocks/.env, BLOCKS_RUN_KEYS; make more on the run keys page.

Address: https://brickbank.cc/b/dibs

Call it:

# Use a new code for each new run. Send the same code again when you retry, and you get the first run back, never charged twice.
curl -X POST https://brickbank.cc/b/dibs/v1/run/names.suggest -H "Authorization: Bearer <YOUR_RUN_KEY>" -H "Idempotency-Key: <NEW_CODE_PER_RUN>" -d '{"items": [{"id": "example", "input": {"vision": "- **V-1** `cap` \u2014 Founders get a landing page critique in one minute.", "count": 5}}]}'

MCP address: https://brickbank.cc/b/dibs/mcp (tools)

Claude Code:

claude mcp add --transport http dibs https://brickbank.cc/b/dibs/mcp --header "Authorization: Bearer <YOUR_RUN_KEY>"

.mcp.json (other MCP clients):

{
  "mcpServers": {
    "dibs": {
      "type": "http",
      "url": "https://brickbank.cc/b/dibs/mcp",
      "headers": {
        "Authorization": "Bearer <YOUR_RUN_KEY>"
      }
    }
  }
}

Run keys page (needs a run key)

naming.internal

Names for an internal tool, best first, each with how it was made and, when generated, the blind guess

Does *

Siblings

Taken

Wraps

Family

Count

Example call

# Use a new code for each new run. Send the same code again when you retry, and you get the first run back, never charged twice.
curl -X POST https://brickbank.cc/b/dibs/v1/run/naming.internal -H "Authorization: Bearer <YOUR_RUN_KEY>" -H "Idempotency-Key: <NEW_CODE_PER_RUN>" -d '{"items": [{"id": "example", "input": {"does": "One read-only view of every brainstorm, spec and story across the repos, and which work went quiet.", "siblings": [{"name": "fleet", "does": "runs many coding-agent workers across every repo under one master"}, {"name": "tower", "does": "the control MCP over the worker scheduler"}], "taken": ["lane", "gate", "board"]}}]}'

smspool 1.0.0

Outside sites it contacts: api.smspool.net

View manifest

sms.code — Rent a phone number for a site and read the code it texts $0.01 per unit Inputs: Site the code is for, Country of the number (ISO-3166 alpha-2), SMSPool pool id; blank lets SMSPool pick the cheapest Try
balance.read — The SMSPool credit left on your key, in USD $0.00 per unit Inputs: none Try
key.validate — Whether an SMSPool API key is valid, free of charge $0.00 per unit Inputs: none Try

From your software

Replace <YOUR_RUN_KEY> with your run key. The operator's run key is in crew /root/dev/blocks/.env, BLOCKS_RUN_KEYS; make more on the run keys page.

Address: https://brickbank.cc/b/smspool

Call it:

# Use a new code for each new run. Send the same code again when you retry, and you get the first run back, never charged twice.
curl -X POST https://brickbank.cc/b/smspool/v1/run/sms.code -H "Authorization: Bearer <YOUR_RUN_KEY>" -H "Idempotency-Key: <NEW_CODE_PER_RUN>" -d '{"items": [{"id": "example", "input": {"service": "linkedin", "country": "US"}}]}'

MCP address: https://brickbank.cc/b/smspool/mcp (tools)

Claude Code:

claude mcp add --transport http smspool https://brickbank.cc/b/smspool/mcp --header "Authorization: Bearer <YOUR_RUN_KEY>"

.mcp.json (other MCP clients):

{
  "mcpServers": {
    "smspool": {
      "type": "http",
      "url": "https://brickbank.cc/b/smspool/mcp",
      "headers": {
        "Authorization": "Bearer <YOUR_RUN_KEY>"
      }
    }
  }
}

Run keys page (needs a run key)

capture 0.1.0

Outside sites it contacts: none declared

View manifest

capture.bundle — Renders a URL or an offline HTML mockup in headless Chromium at the given widths and declared states and returns one hashed evidence bundle: full-page and CSS-stripped wireframe PNGs, DOM, accessibility tree, computed styles with one selector per visible element, axe-core results, a bounded click graph and scripted traces. Same input, same bundle_sha. $0.01 per unit Inputs: url, html, assets, widths, states, hide, clip, color_scheme, tasks Try

From your software

Replace <YOUR_RUN_KEY> with your run key. The operator's run key is in crew /root/dev/blocks/.env, BLOCKS_RUN_KEYS; make more on the run keys page.

Address: https://brickbank.cc/b/capture

capture.bundle has no example input yet, so there is no ready-made call: try it to see its inputs.

MCP address: https://brickbank.cc/b/capture/mcp (tools)

Claude Code:

claude mcp add --transport http capture https://brickbank.cc/b/capture/mcp --header "Authorization: Bearer <YOUR_RUN_KEY>"

.mcp.json (other MCP clients):

{
  "mcpServers": {
    "capture": {
      "type": "http",
      "url": "https://brickbank.cc/b/capture/mcp",
      "headers": {
        "Authorization": "Bearer <YOUR_RUN_KEY>"
      }
    }
  }
}

Run keys page (needs a run key)

ui-critic 0.1.0

Outside sites it contacts: api.anthropic.com

Recall not measured · False alarms not measured · Gate-grade no (last 30 days)

View manifest

ui.critique — Judges how a captured screen looks, from its pixels, computed styles and, when you give them, your design tokens only: nine deterministic floor rules (contrast, unstyled controls, tap targets, 390 px overflow, token and grid drift, type sizes, axe, dead clicks) plus a three-sample model quorum over real and wireframe tiles. Returns a critic-verdict/1 report: exit code, admissible findings with locators and crops, JUnit, SARIF and a builder view. Never sees the screen's goal or page source. $0.80 per unit Inputs: stage, kind, bundle, states, rubric, tokens, references, baseline, changed_regions, sanctioned, shadow, builder_session, builder_model, color_scheme Try

From your software

Replace <YOUR_RUN_KEY> with your run key. The operator's run key is in crew /root/dev/blocks/.env, BLOCKS_RUN_KEYS; make more on the run keys page.

Address: https://brickbank.cc/b/ui-critic

ui.critique has no example input yet, so there is no ready-made call: try it to see its inputs.

MCP address: https://brickbank.cc/b/ui-critic/mcp (tools)

Claude Code:

claude mcp add --transport http ui-critic https://brickbank.cc/b/ui-critic/mcp --header "Authorization: Bearer <YOUR_RUN_KEY>"

.mcp.json (other MCP clients):

{
  "mcpServers": {
    "ui-critic": {
      "type": "http",
      "url": "https://brickbank.cc/b/ui-critic/mcp",
      "headers": {
        "Authorization": "Bearer <YOUR_RUN_KEY>"
      }
    }
  }
}

Run keys page (needs a run key)

ux-critic 0.1.0

Outside sites it contacts: api.anthropic.com

Recall not measured · False alarms not measured · Gate-grade no (last 30 days)

View manifest

ux.critique — Judges whether a persona can finish their tasks on a screen, from a capture bundle's accessibility trees and click graph (never its pixels), from a written UX spec, or from an API/library example script and its output. Deterministic graph rules plus a 3-sample LLM walk quorum; returns a critic-verdict/1 report (exit code, findings, JUnit, SARIF, builder view) and the walked traces. $0.85 per unit Inputs: stage, kind, bundle, ux_spec, dx, persona, tasks, states, rubric, baseline, changed_regions, sanctioned, shadow, builder_session, builder_model Try

From your software

Replace <YOUR_RUN_KEY> with your run key. The operator's run key is in crew /root/dev/blocks/.env, BLOCKS_RUN_KEYS; make more on the run keys page.

Address: https://brickbank.cc/b/ux-critic

ux.critique has no example input yet, so there is no ready-made call: try it to see its inputs.

MCP address: https://brickbank.cc/b/ux-critic/mcp (tools)

Claude Code:

claude mcp add --transport http ux-critic https://brickbank.cc/b/ux-critic/mcp --header "Authorization: Bearer <YOUR_RUN_KEY>"

.mcp.json (other MCP clients):

{
  "mcpServers": {
    "ux-critic": {
      "type": "http",
      "url": "https://brickbank.cc/b/ux-critic/mcp",
      "headers": {
        "Authorization": "Bearer <YOUR_RUN_KEY>"
      }
    }
  }
}

Run keys page (needs a run key)

belief-extractor 0.1.0

Outside sites it contacts: none declared

View manifest

company.beliefs — The beliefs a company states, found in its marketing pages, its job posts and the industry content around it. One result per belief, with the quotes that state it and the challenge it implies. $0.00 per unit Inputs: Company website Try
company.internal_voice — What a company says about how it hires, its hiring bar and its culture, in its own words: its careers page and its open job posts, quoted word for word under one header per source. A source that failed is named with its reason, never left out. $0.00 per unit Inputs: Company website Try
company.founder_profiles — Who a company's founders and chief executives are on LinkedIn: up to 3 profile URLs, found from the LinkedIn page the company's website links and kept only when their current title says founder, CEO, CTO or another chief. A company with none is a miss naming why. $0.04 per unit Inputs: Company website Try
company.founder_posts — What people say in public, in their own words: the LinkedIn posts (not reposts) of up to 3 profiles from the last six months, at most 3 each, quoted word for word under one header per post. Give it company.founder_profiles' answer to read a company's founders. A source that failed is named with its reason, never left out. $0.04 per unit Inputs: Founders' LinkedIn profiles Try
company.founder_background — Where people have worked and studied, in their own words: the experience and education of up to 3 LinkedIn profiles, quoted word for word under one header per person. It answers whether a founder is a domain expert (a clinician, a lawyer, a teacher) and not only a software engineer. Give it company.founder_profiles' answer to read a company's founders. A source that failed is named with its reason, never left out. $0.01 per unit Inputs: Founders' LinkedIn profiles Try

From your software

Replace <YOUR_RUN_KEY> with your run key. The operator's run key is in crew /root/dev/blocks/.env, BLOCKS_RUN_KEYS; make more on the run keys page.

Address: https://brickbank.cc/b/belief-extractor

Call it:

# Use a new code for each new run. Send the same code again when you retry, and you get the first run back, never charged twice.
curl -X POST https://brickbank.cc/b/belief-extractor/v1/run/company.beliefs -H "Authorization: Bearer <YOUR_RUN_KEY>" -H "Idempotency-Key: <NEW_CODE_PER_RUN>" -d '{"items": [{"id": "example", "input": {"company_domain": "jackandjill.ai"}}]}'

MCP address: https://brickbank.cc/b/belief-extractor/mcp (tools)

Claude Code:

claude mcp add --transport http belief-extractor https://brickbank.cc/b/belief-extractor/mcp --header "Authorization: Bearer <YOUR_RUN_KEY>"

.mcp.json (other MCP clients):

{
  "mcpServers": {
    "belief-extractor": {
      "type": "http",
      "url": "https://brickbank.cc/b/belief-extractor/mcp",
      "headers": {
        "Authorization": "Bearer <YOUR_RUN_KEY>"
      }
    }
  }
}

Run keys page (needs a run key)

company.founder_background

Where people have worked and studied, in their own words: the experience and education of up to 3 LinkedIn profiles, quoted word for word under one header per person. It answers whether a founder is a domain expert (a clinician, a lawyer, a teacher) and not only a software engineer. Give it company.founder_profiles' answer to read a company's founders. A source that failed is named with its reason, never left out.

Founders' LinkedIn profiles *

Example call

# Use a new code for each new run. Send the same code again when you retry, and you get the first run back, never charged twice.
curl -X POST https://brickbank.cc/b/belief-extractor/v1/run/company.founder_background -H "Authorization: Bearer <YOUR_RUN_KEY>" -H "Idempotency-Key: <NEW_CODE_PER_RUN>" -d '{"items": [{"id": "example", "input": {"founder_profile_urls": ["https://www.linkedin.com/in/williamhgates"]}}]}'

One block. Three ways to run it.

Same block, same price, wherever you run it.

This page Fill the form, check the price per result, press Run.

Software One API call. Quick jobs answer at once, long ones send a receipt.

AI agents Claude and other assistants use it as a tool, inside your budget.

Brickbank· Finished jobs, paid per result.

Finished jobs, paid per result.

Blocks you can run today.

appsim 0.2.0

read_job_ad

read_resume

match_report

bossroster 0.1.0

resolve

aggregate

leader_seat

people.find

ask

email.find

key.validate

lpcritic 0.1.0

line_check

hero_check

page_critique

instantly 1.0.0

campaign.enrol

lead.add

lead.list

campaign.sending_status

lemlist 1.0.0

campaign.enrol

campaigns.list

campaigns.inventory

campaign.preflight

lead.add

activities.pull

seats.count

key.validate

employment.confirm

people.stated_employment

whatblock 0.1.0

company.what_it_does

datagate-people 0.1.0

bettercontact.email.find

bettercontact.phone.find

bettercontact.people.find

bettercontact.key.validate

bossroster.people.find

bossroster.email.find

bossroster.key.validate

companies_house.companies.search

companies_house.filings.share_allotments

companies_house.officers.list

companies_house.key.validate

getprospect.email.find

getprospect.email.verify

getprospect.key.validate

hunter.email.find

hunter.email.find_published

hunter.email.verify

hunter.people.find

hunter.key.validate

hunter.account.balance

icypeas.people.rank

icypeas.people.taxonomy

icypeas.domain.scan

icypeas.email.find

icypeas.email.verify

icypeas.account.balance

icypeas.companies.search

icypeas.company.find

icypeas.key.validate

icypeas.people.count

icypeas.people.find

icypeas.people.find_ranked

icypeas.people.search

icypeas.company.scrape

icypeas.company.url_find

icypeas.email.reverse

icypeas.profile.scrape

icypeas.profile.url_find

krs.officers.list

krs.connection.try

leadmagic.email.find

leadmagic.people.find

leadmagic.phone.find