toolbelt402.tpoborne.workers.dev publishes 20 machine-payable endpoints over the x402 protocol, priced from $0.001 to $0.02 per call. Each was probed directly; the response code below is what it returned.
| Endpoint | Method | Price | Probe |
|---|---|---|---|
| /cron/next | GET | $0.001 | 200 |
| /csv/to-json | POST | $0.003 | 200 |
| /encode | POST | $0.001 | 200 |
| /hash | POST | $0.002 | 200 |
| /html/to-markdown | POST | $0.005 | 200 |
| /json/diff | POST | $0.003 | 200 |
| /json/query | POST | $0.003 | 200 |
| /json/repair | POST | $0.005 | 200 |
| /json/to-csv | POST | $0.002 | 200 |
| /json/validate | POST | $0.002 | 402 |
| /jwt/decode | POST | $0.002 | 200 |
| /markdown/to-html | POST | $0.003 | 200 |
| /regex/test | POST | $0.002 | 200 |
| /text/diff | POST | $0.002 | 200 |
| /time/math | GET | $0.001 | 200 |
| /url/parse | GET | $0.001 | 200 |
| /url/preflight | GET | $0.005 | 402 |
| /validate/checksum | POST | $0.002 | 200 |
| /web/extract | POST | $0.02 | 200 |
| /xml/to-json | POST | $0.003 | 200 |
Parse and validate a cron expression — returns the next N run times as ISO timestamps in any IANA timezone. Cron schedule checker.
Parse CSV into JSON rows — auto-detects the delimiter, handles quoted fields and headers, optional type coercion. Returns columns, row count, and per-row parse errors.
Encode or decode text: base64, base64url (padding handled), hex, URL percent-encoding, and HTML entities. The conversions agents most often get wrong.
Hash a string or bytes with multiple algorithms in one call: SHA-1, SHA-256, SHA-384, SHA-512, SHA3-256, and keccak-256 (Ethereum). Input as UTF-8, base64, or hex; hex digests out.
Convert HTML to clean Markdown — turn scraped web pages into LLM-ready text. Handles headings, lists, tables, links, code blocks. Deterministic, no LLM.
Structural diff of two JSON documents — add/remove/change operations with JSON-Pointer paths, before/after values, and an identical flag. Compare API responses and configs.
Query JSON with a JMESPath expression — extract, filter, project, and reshape nested data with the AWS-standard query language. Deterministic JSON extraction.
Fix invalid, malformed, or truncated JSON — repairs trailing commas, single quotes, unquoted keys, and cut-off LLM output; returns the parsed object plus corrected JSON text. Deterministic, no LLM.
Convert an array of JSON objects into CSV — column union across rows or explicit column order; nested values are JSON-stringified into cells.
Validate JSON data against a JSON Schema (draft-07, 2019-09, 2020-12) — returns exact error locations, keywords, and messages for programmatic handling. Deterministic schema validation.
Decode a JWT — header, payload, and timing analysis (issued/not-before/expiry as ISO timestamps, expired flag, seconds until expiry). Decode only; signature is NOT verified.
Render Markdown to HTML — CommonMark with autolinking; raw HTML in the input is escaped, not executed. The inverse of /html/to-markdown.
Test and debug a regular expression against up to 50 sample strings — real match results, capture groups, named groups, and pass/fail against expectations. Deterministic regex testing.
Compare two texts — unified diff patch plus added/removed/unchanged counts, by lines, words, or characters. Verify edits and detect changes deterministically.
Date and time arithmetic done right: difference between two datetimes, add or subtract ISO-8601 durations, convert an instant into any IANA timezone with UTC offset and weekday.
Parse a URL into parts — host, path, query params as JSON, fragment, origin — or resolve a relative reference against a base URL correctly.
Check a URL before fetching it: is it alive, final URL after redirects, HTTP status, content type, HTML title, and whether robots.txt allows your user-agent. Link checker and fetch preflight for agent…
Validate identifiers by checksum: IBAN (mod-97), card numbers (Luhn), EAN-13, ISBN-10/13, UUID (RFC-4122), ULID. Batch up to 100 per call with per-item reasons.
Read a web page for me: fetches up to 5 URLs and returns the main article content as clean Markdown (nav, ads, and footers stripped) plus title, author, published date, canonical URL, links, images, w…
Parse XML into JSON — validates first with line-numbered errors, preserves attributes, optional type coercion. Works for RSS feeds, sitemaps, and SOAP-ish payloads.