"The official Python standard library documentation" matching MCP connectors:
GET /v1/connectors – MCP directory API referenceMatching Connector Tools:
Test inboxes for your app's sign-up and login emails: get the code or link, or why none came.
Verify before the agent pays. Screwtape is an independent check on agent work before a payment goes out: invoices, duplicate invoices and reconciliation, with a signed receipt for your auditor. It is independent of your agent's own model and vendor, with no stake in your agent passing. - judge: pass, fail or retry verdict on one agent output, with a score, reasons and a next_step. Optional deterministic checks: required fields, types, formats, sums, regex. - check_duplicate_invoice (/v1/duplicate): opt-in duplicate-invoice check using keyed fingerprints. - consensus: scores agreement across 2 to 16 agent answers and flags false consensus. - Advisory mode: recommends approve, review or return; a person decides. - Receipts: Ed25519-signed. The receipt proves what was checked, when, and the verdict. Verify via POST /v1/receipts/verify or offline; export via GET /v1/receipts/export. - Judged content isn't stored unless you opt in. Free key: 100 calls/day. Or $0.005 USDC per call via x402 on Base or Solana, no key needed. MCP: https://screwtape.ai/mcp Docs: https://screwtape.ai/docs/ Icon: https://screwtape.ai/icon-512.png Tags: verification, evaluation, llm-as-judge, invoices, receipts, x402
See which AI models fall for traps in everyday jobs, and check or scan any MCP server, skill or package before your agent installs it. No sign-in; read-only; returns structured data.
The Decision Operations Platform for engineering teams
Disposable, receive-only email inboxes for AI agents. Create an address, wait for the verification email, and get the one-time code or magic link back as structured fields. Long-polling wait_for_email with sender and subject filters, attachments, sanitised HTML, and restore for later password resets. OAuth 2.1, hosted in the EU, cannot send email.
Official Mifori Agent Trust Protocol. Performs real-time latency, JSON schema compliance, and synthetic IQ audits on target AI bots using x402 USDC micro-payments on Base network before high-value tool execution.
Benchmark your own prompt on 100+ models with real API calls, then choose the winner by accuracy, cost, speed, or any mix of those.
Test your AI agents for security and safety weaknesses, and re-verify the fixes.
Validate JSON-LD structured data against the Schema.org vocabulary. Free; never fetches URLs.
Security scans for AI-built web apps, starting with Lovable. Verify ownership, run authorized scans, inspect findings, get AI-ready fix prompts and request retests. Uses OAuth with a Guarita account. Fix prompt access follows the account plan.
Read browser bug recordings: video, console, network, steps. Fix the bug and comment back.
Fork-replay one recorded span from a failed agent run and get an attribution verdict.
You cancelled the run — did the work actually stop?
Runs an eight-stage isolation test against your staging Supabase and issues a signed attestation. Needs a RowAttest API key: Authorization: Bearer rak_... (create one on the API keys page in the app).
Review supplied AI claims and source excerpts for quote mismatches, broken references and evidence gaps. MCP-native free_trial_status/free_trial: five attempts per network for the campaign, shared daily capacity. Paid checks are separate. No live retrieval or factual-truth certification; human review required.
Pair-level MCP+A2A conformance verdicts for the exact pair you delegate across.
Website monitoring for AI assistants: catches pages that look broken even when the site is up.
Validate and clean CSV before import. Find duplicates, missing values and invalid dates with row-level reports. Apply only the cleanup you request; files are not retained. Run view_csv_demo free without a key. Custom CSV: EUR 9 for 100 operations, valid 90 days, no subscription. Setup: https://check.orvel.dev/docs/
Pre-run parity probe for scheduled agents: signed PARITY / NO-GO verdict naming the failing gate.
Rehearsal range: mirror a site, rehearse the workflow, earn a drift-bound promotion receipt.
