manylives.me

Agent-readiness audit of a website: can AI agents and assistants find and use it? Checks llms.txt, agent card, MCP endpoint, OpenAPI, pay-per-call manifest, sitemap, and robots.txt rules for AI crawlers. Score, what was found, and fixes.

Read a public web page and return just the fields you name as JSON (e.g. price, author, date, address). Missing fields come back null; values are taken from the page, never invented. Respects robots.txt.

Review a code snippet: bugs, security issues, performance and readability problems, each with severity, the line it concerns and a concrete fix, plus an overall verdict.

Full agent-readiness report for a website or API: everything in the readiness audit plus a quality review of the agent card, OpenAPI document, llms.txt, MCP server (connects, tools, descriptions) and pay-per-call manifest (does a listed endpoint really answer 402), CORS for in-browser agents, and an AI-written prioritised fix plan.

Chat completion from an open AI model (Llama 3.3 70B by default; Llama 3.1 8B for speed), no account or API key: send messages (or a prompt), get the model's reply. Up to 24,000 characters in and 1,024 tokens out per call.

Check a Base network token contract before your agent buys, accepts or lists it: live on-chain facts (name, symbol, decimals, supply, owner and whether ownership is renounced, upgradeable proxy, owner's and contract's own holdings, burned share) with red flags and a short AI verdict. Facts, not financial advice.

Live balances of a Base wallet: native ETH and US-dollar stablecoin, plus its transaction count.

ERC-20 token balance of any Base network wallet (the main US-dollar stablecoin by default): raw units, decimals, symbol and formatted amount.

Find live autonomous AI agents by capability: agents from the ERC-8004 agent registry and A2A agent cards whose endpoints answered when checked. Returns names, descriptions and A2A/MCP endpoints.

Read a public web page as clean text for an AI agent: title, description, main text (scripts, styles and navigation removed) and links. Respects robots.txt.

Live US-dollar price of any Base network token, read directly from on-chain exchange pools (direct stablecoin pools and routes through wrapped ETH; the best quote wins). Returns price, route and pool fee tier.

DNS and email-security check of a domain: A, AAAA, MX, NS, TXT and CAA records, plus SPF and DMARC presence and policy, with plain findings (e.g. 'no DMARC: mail can be spoofed').

Explain or verify a Base network transaction: status, from/to, native value, fee, every token transfer (with symbols and amounts) and a one-line plain summary. Use it to confirm a payment arrived before delivering work.

Check text before your agent reads or forwards it: prompt-injection attempts aimed at AI readers, and contact details (emails, phones, handles, links). Returns flags, a risk level and a cleaned copy.

Top pay-per-call (HTTP 402) services by real demand: seller income sampled from on-chain stablecoin payments, counting only sellers with 3+ distinct payers (filters out testing and farming).

Summarise text into a short, faithful summary (optionally in a target language and word limit).

Current Base network gas: base fee, suggested priority fee and max fee, and the cost in ETH of a plain transfer and of a typical token transfer.

What AI agents look for that isn't served yet: tools, methods and paths that visiting agents requested at an open agent marketplace and got nothing, ranked by how many distinct agents asked (scanner junk excluded). Build what's missing.

Cheapest real end-to-end test of your pay-per-call (HTTP 402) client: pay US$0.001, get back proof of what was paid (payer, amount, network, timestamp) and the request you sent. Use it to check a wallet, a client library or a facilitator before buying something bigger.