Skip to content
Closed
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
34 changes: 31 additions & 3 deletions src/adapters/responses-tool-schema.ts
Original file line number Diff line number Diff line change
Expand Up @@ -96,9 +96,32 @@ function usesUnicodePropertyEscape(pattern: string): boolean {
return false;
}

/**
* Rewrite each unescaped `\0` not followed by a digit to the equivalent `\x00`. Both spell NUL
* in ECMAScript and Python `re`, but some destinations' schema validators refuse `\0` inside a
* character class: Meta's Responses API answers 400 "is not a \"regex\"" to Claude Code's
* Artifact tool, whose file-path parameters carry `^[^\0]*$`. `\0` followed by a digit is an
* octal escape in Python, so it is left alone. Returns the input when nothing changed.
*/
function rewriteNulEscapes(pattern: string): string {
let out: string | undefined;
let copied = 0;
for (let i = 0; i < pattern.length; i++) {
if (pattern[i] !== "\\") continue;
const next = pattern[i + 1];
if (next === "0" && !/[0-9]/.test(pattern[i + 2] ?? "")) {
out = (out ?? "") + pattern.slice(copied, i) + "\\x00";
copied = i + 2;
}
i++;
}
return out === undefined ? pattern : out + pattern.slice(copied);
}

/**
* Remove unsupported Unicode property escapes from scalar `pattern` constraints in ordinary
* positive schema positions. This keeps built-in Artifact tools usable on Python-re backends;
* positive schema positions, and spell `\0` as `\x00` (rewriteNulEscapes). This keeps built-in
* Artifact tools usable on Python-re backends and on Meta's Responses API;
* the omitted constraint is not enforced by this proxy and tools must validate their inputs.
*
* Regex-keyed objects are preserved. Removing a matcher can lose evaluated-property annotations
Expand Down Expand Up @@ -181,8 +204,13 @@ export function stripUnicodePropertyPatterns(node: unknown, inNameBag = false):
continue;
}
const [key, value] = next.value;
if (!frame.inNameBag && key === "pattern" && typeof value === "string" && usesUnicodePropertyEscape(value)) {
delete (cloneContainer(frame) as Record<string, unknown>)[key];
if (!frame.inNameBag && key === "pattern" && typeof value === "string") {
if (usesUnicodePropertyEscape(value)) {
delete (cloneContainer(frame) as Record<string, unknown>)[key];
continue;
}
const rewritten = rewriteNulEscapes(value);
if (rewritten !== value) (cloneContainer(frame) as Record<string, unknown>)[key] = rewritten;
continue;
}
if (!frame.inNameBag && (PRESERVED_PATTERN_SUBTREES.has(key) || SCHEMA_LITERAL_VALUE_KEYS.has(key))) {
Expand Down
23 changes: 23 additions & 0 deletions tests/adapters/openai/openai-chat-hardening.test.ts
Original file line number Diff line number Diff line change
Expand Up @@ -362,6 +362,29 @@ describe("unicode property-escape pattern stripping", () => {
expect(stripUnicodePropertyPatterns(before)).toBe(before);
});

test("`\\0` is rewritten to the equivalent `\\x00`, octal and escaped backslashes untouched", () => {
// Claude Code's Artifact tool ships `^[^\0]*$` on its file-path parameters; Meta's
// Responses API rejects `\0` inside a character class but accepts `\x00`.
const artifactPathPattern = "^[^\\0]*$";
const before = {
type: "object",
properties: {
path: { type: "string", pattern: artifactPathPattern, maxLength: 1024 },
octal: { type: "string", pattern: "^\\012$" },
literal: { type: "string", pattern: "^\\\\0$" },
},
};
const out = stripUnicodePropertyPatterns(before) as typeof before;
expect(out.properties.path.pattern).toBe("^[^\\x00]*$");
expect(out.properties.path.maxLength).toBe(1024);
expect(out.properties.octal).toBe(before.properties.octal);
expect(out.properties.literal).toBe(before.properties.literal);
expect(before.properties.path.pattern).toBe(artifactPathPattern);
for (const s of ["\u0000", "a", "\u0000/b"]) {
expect(new RegExp(out.properties.path.pattern).test(s)).toBe(new RegExp(artifactPathPattern).test(s));
}
});

test("`\\P{…}` is dropped as well as `\\p{…}`", () => {
const stripped = stripUnicodePropertyPatterns({ type: "string", pattern: "^\\P{L}+$" }) as Record<string, unknown>;
expect(stripped.pattern).toBeUndefined();
Expand Down
Loading