p1: cap jq output, log refusals, fail contract without SDK

jq results are bounded (10k items / output cap) and return 400 when
exceeded. 401/403 log caller, host, and reason. make contract exits 1
unless firecrawl-py is present (SIFT_CONTRACT_SKIP_SDK=1 to skip).

internal/clean/clean.go

@@ -9,6 +9,7 @@ import (
 	"net/url"
 	"strings"
 	"time"
+	"unicode/utf8"
 
 	readability "github.com/go-shiori/go-readability"
 	htmltomarkdown "github.com/JohannesKaufmann/html-to-markdown/v2"
@@ -140,6 +141,7 @@ func JSON(res *fetch.Resource, jqFilter string, outCap int) (*Result, error) {
 		}
 		iter := code.RunWithContext(ctx, v)
 		var acc []any
+		var running int
 		for {
 			x, ok := iter.Next()
 			if !ok {
@@ -149,6 +151,17 @@ func JSON(res *fetch.Resource, jqFilter string, outCap int) (*Result, error) {
 				return nil, fmt.Errorf("jq: %w", err)
 			}
 			acc = append(acc, x)
+			if len(acc) > 10_000 {
+				return nil, &TooLargeError{Msg: "jq output too large"}
+			}
+			b, err := json.Marshal(x)
+			if err != nil {
+				return nil, fmt.Errorf("jq: %w", err)
+			}
+			running += len(b)
+			if running > outCap {
+				return nil, &TooLargeError{Msg: "jq output too large"}
+			}
 		}
 		switch len(acc) {
 		case 0:
@@ -183,15 +196,33 @@ func Text(res *fetch.Resource, outCap int) (*Result, error) {
 	return capResult(out, outCap), nil
 }
 
+// TooLargeError is a jq result that exceeded item or byte caps.
+type TooLargeError struct{ Msg string }
+
+func (e *TooLargeError) Error() string { return e.Msg }
+
 func capResult(r *Result, capn int) *Result {
 	if len(r.Markdown) > capn {
-		r.Markdown = r.Markdown[:capn] + "\n…[truncated]"
+		r.Markdown = TruncateUTF8(r.Markdown, capn) + "\n…[truncated]"
 		r.Trunc = true
 		r.Warning = "truncated to output cap"
 	}
 	return r
 }
 
+func TruncateUTF8(s string, n int) string {
+	if n >= len(s) {
+		return s
+	}
+	if n < 0 {
+		return ""
+	}
+	for n > 0 && !utf8.RuneStart(s[n]) {
+		n--
+	}
+	return s[:n]
+}
+
 func stripDanger(s string) string {
 	doc, err := html.Parse(strings.NewReader(s))
 	if err != nil {