catalog / engines

JSON Parser

Parse and serialise JSON per RFC 8259, with useful errors.

15 capabilities 1 sections reference specification

Build against it

# give your agent the catalog — no account, no install
claude mcp add provespec -- npx -y provespec-mcp

# then, from any agent session
spec_get(slug: "json-parser")                → the capability tree
spec_checklist(slug: "json-parser")          → the questions to answer
spec_grade(slug: "json-parser", grades: [...])  → parity % + the gap list

# or just fetch it
curl https://provespec.com/catalog/json-parser/spec.json

Grade every capability below as yes · partial · no · na, with a file reference as evidence. Whatever you leave ungraded counts as missing — the gap list is the work queue.

Or just paste this to your agent

Use the ProveSpec MCP server. Fetch the "json-parser" spec with
spec_get, then walk spec_checklist and grade this repository against it —
one verdict per capability, each with a file reference as the note.
Call spec_grade with the results and give me the gap list as the plan.

If the server is not registered, add it first:
  claude mcp add provespec -- npx -y provespec-mcp

The specification

AI 15

Parse
  • Objects, arrays, strings, numbers, booleans, null the whole grammar
  • String escapes including \u surrogate pairs
  • Numbers with exponent and fraction no leading zeros, no hex
  • Reject trailing commas and comments unless a relaxed mode is opted into
  • Reject duplicate-key ambiguity explicitly documented behaviour either way
  • Nesting depth limit guards against stack exhaustion
Errors
  • Message names the byte offset, line and column
  • Errors say what was expected not just "unexpected token"
Serialise
  • Round-trip parse and stringify value equality preserved
  • Pretty-print with configurable indent
  • Stable key ordering option for reproducible output
  • Escape control characters and optionally non-ASCII
Streaming
  • Incremental parse of large documents without holding the whole text
  • Event or token stream API
Correctness
  • Passes the JSONTestSuite corpus y_/n_/i_ cases classified
Download spec.json All 53 specs