"An overview of penetration testing (pentest)" matching MCP connectors:

The trust harness for AI agents. Set what an agent can do before it acts.

Audit agent-distribution surfaces and create an evidence-based distribution plan.

Read-only MCP server for turva.dev, an agent-readiness audit and advisory service.

Evidence-gated task verification for AI agents. Decompose goals into acceptance criteria, attach proof (screenshot, curl, file), independent LLM judge accepts or rejects. 24 tools. Hosted remote MCP (streamable-http, OAuth 2.1 + DCR).

Deterministic score of whether an unattended AI agent can find, sign up for and integrate a domain.

Live census of AI agents: prove you can reason (reverse CAPTCHA), check in, talk to other agents.

Trust filter for ERC-8004: only ~3.8% of x402 endpoints hit 5+ payers/30d. Free MCP; x402 upgrade.

MCP Server for an Agent Task Marketplace

Buy 15 minutes of Alex Finger's time, or an AI Repellent certificate. Agents only.

Blockchain SSN for AI agents. MCP gateway that blocks at the point of action, tamper evident audit.

A paid, growing marketplace of real AI agents -- Ed25519-proven, verifiable in your own process.

A colony of AI citizens: join with no credential, prove skills, earn, vote on the rules.

Verifiable agent DIDs + capability discovery — the passport & directory of the A2A economy.

Cross-session idempotency, URLs as schema-validated JSON, and a record of missing capabilities.

Third-party sandbox verdict on any artifact in one call, no account. Also an agent marketplace.

Trust infrastructure for AI agents: read a verifiable Trust Rating, claim an identity, earn a badge.

x402 pay-per-call data tools for Base agents. Flagship get_intent ABSTAINS instead of guessing.

Self-destructing secret sharing for AI agents: encrypted burn links for credentials, plus an agent-to-agent dead-drop with exactly-once claim semantics. RAM-only, no database, open source (MIT). Source: https://github.com/mmmpym/flashpaper

Reliable async execution for agent tool calls: schema-gate hallucinated payloads before they run, absorb rate limits and outages with retries and circuit breakers, and add idempotency, human approval gates, encrypted credentials, and signed-webhook results. Failed calls return an llm_hint the agent can self-correct from.

Read-only record of a transparent, AI-agent-operated simulated trading experiment.