Files
openclaw/src/agents/pi-embedded-helpers.sanitizeuserfacingtext.e2e.test.ts
niceysam 5e423b596c fix: remove false-positive billing error rewrite on normal assistant text (openclaw#17834) thanks @niceysam
Verified:
- pnpm install --frozen-lockfile
- pnpm build
- pnpm check
- pnpm test:macmini

Co-authored-by: niceysam <256747835+niceysam@users.noreply.github.com>
Co-authored-by: Tak Hoffman <781889+Takhoffman@users.noreply.github.com>
2026-02-21 12:17:39 -06:00

415 lines
14 KiB
TypeScript

import { describe, expect, it } from "vitest";
import {
downgradeOpenAIReasoningBlocks,
isMessagingToolDuplicate,
normalizeTextForComparison,
sanitizeToolCallId,
sanitizeUserFacingText,
stripThoughtSignatures,
} from "./pi-embedded-helpers.js";
describe("sanitizeUserFacingText", () => {
it("strips final tags", () => {
expect(sanitizeUserFacingText("<final>Hello</final>")).toBe("Hello");
expect(sanitizeUserFacingText("Hi <final>there</final>!")).toBe("Hi there!");
});
it("does not clobber normal numeric prefixes", () => {
expect(sanitizeUserFacingText("202 results found")).toBe("202 results found");
expect(sanitizeUserFacingText("400 days left")).toBe("400 days left");
});
it("sanitizes role ordering errors", () => {
const result = sanitizeUserFacingText("400 Incorrect role information", { errorContext: true });
expect(result).toContain("Message ordering conflict");
});
it("sanitizes HTTP status errors with error hints", () => {
expect(sanitizeUserFacingText("500 Internal Server Error", { errorContext: true })).toBe(
"HTTP 500: Internal Server Error",
);
});
it("sanitizes direct context-overflow errors", () => {
expect(
sanitizeUserFacingText(
"Context overflow: prompt too large for the model. Try /reset (or /new) to start a fresh session, or use a larger-context model.",
{ errorContext: true },
),
).toContain("Context overflow: prompt too large for the model.");
expect(
sanitizeUserFacingText("Request size exceeds model context window", { errorContext: true }),
).toContain("Context overflow: prompt too large for the model.");
});
it("does not swallow assistant text that quotes the canonical context-overflow string", () => {
const text =
"Changelog note: we fixed false positives for `Context overflow: prompt too large for the model. Try /reset (or /new) to start a fresh session, or use a larger-context model.` in 2026.2.9";
expect(sanitizeUserFacingText(text)).toBe(text);
});
it("does not rewrite conversational mentions of context overflow", () => {
const text =
"nah it failed, hit a context overflow. the prompt was too large for the model. want me to retry it with a different approach?";
expect(sanitizeUserFacingText(text)).toBe(text);
});
it("does not rewrite technical summaries that mention context overflow", () => {
const text =
"Problem: When a subagent reads a very large file, it can exceed the model context window. Auto-compaction cannot help in that case.";
expect(sanitizeUserFacingText(text)).toBe(text);
});
it("does not rewrite conversational billing/help text without errorContext", () => {
const text =
"If your API billing is low, top up credits in your provider dashboard and retry payment verification.";
expect(sanitizeUserFacingText(text)).toBe(text);
});
it("does not rewrite normal text that mentions billing and plan", () => {
const text =
"Firebase downgraded us to the free Spark plan; check whether we need to re-enable billing.";
expect(sanitizeUserFacingText(text)).toBe(text);
});
it("does not rewrite billing error-shaped text without errorContext", () => {
const text = "billing: please upgrade your plan";
expect(sanitizeUserFacingText(text)).toBe(text);
});
it("rewrites billing error-shaped text with errorContext", () => {
const text = "billing: please upgrade your plan";
expect(sanitizeUserFacingText(text, { errorContext: true })).toContain("billing error");
});
it("sanitizes raw API error payloads", () => {
const raw = '{"type":"error","error":{"message":"Something exploded","type":"server_error"}}';
expect(sanitizeUserFacingText(raw, { errorContext: true })).toBe(
"LLM error server_error: Something exploded",
);
});
it("returns a friendly message for rate limit errors in Error: prefixed payloads", () => {
expect(sanitizeUserFacingText("Error: 429 Rate limit exceeded", { errorContext: true })).toBe(
"⚠️ API rate limit reached. Please try again later.",
);
});
it("collapses consecutive duplicate paragraphs", () => {
const text = "Hello there!\n\nHello there!";
expect(sanitizeUserFacingText(text)).toBe("Hello there!");
});
it("does not collapse distinct paragraphs", () => {
const text = "Hello there!\n\nDifferent line.";
expect(sanitizeUserFacingText(text)).toBe(text);
});
it("strips leading newlines from LLM output", () => {
expect(sanitizeUserFacingText("\n\nHello there!")).toBe("Hello there!");
expect(sanitizeUserFacingText("\nHello there!")).toBe("Hello there!");
expect(sanitizeUserFacingText("\n\n\nMultiple newlines")).toBe("Multiple newlines");
});
it("strips leading whitespace and newlines combined", () => {
expect(sanitizeUserFacingText("\n \nHello")).toBe("Hello");
expect(sanitizeUserFacingText(" \n\nHello")).toBe("Hello");
});
it("preserves trailing whitespace and internal newlines", () => {
expect(sanitizeUserFacingText("Hello\n\nWorld\n")).toBe("Hello\n\nWorld\n");
expect(sanitizeUserFacingText("Line 1\nLine 2")).toBe("Line 1\nLine 2");
});
it("returns empty for whitespace-only input", () => {
expect(sanitizeUserFacingText("\n\n")).toBe("");
expect(sanitizeUserFacingText(" \n ")).toBe("");
});
});
describe("stripThoughtSignatures", () => {
it("returns non-array content unchanged", () => {
expect(stripThoughtSignatures("hello")).toBe("hello");
expect(stripThoughtSignatures(null)).toBe(null);
expect(stripThoughtSignatures(undefined)).toBe(undefined);
expect(stripThoughtSignatures(123)).toBe(123);
});
it("removes msg_-prefixed thought_signature from content blocks", () => {
const input = [
{ type: "text", text: "hello", thought_signature: "msg_abc123" },
{ type: "thinking", thinking: "test", thought_signature: "AQID" },
];
const result = stripThoughtSignatures(input);
expect(result).toHaveLength(2);
expect(result[0]).toEqual({ type: "text", text: "hello" });
expect(result[1]).toEqual({
type: "thinking",
thinking: "test",
thought_signature: "AQID",
});
expect("thought_signature" in result[0]).toBe(false);
expect("thought_signature" in result[1]).toBe(true);
});
it("preserves blocks without thought_signature", () => {
const input = [
{ type: "text", text: "hello" },
{ type: "toolCall", id: "call_1", name: "read", arguments: {} },
];
const result = stripThoughtSignatures(input);
expect(result).toEqual(input);
});
it("handles mixed blocks with and without thought_signature", () => {
const input = [
{ type: "text", text: "hello", thought_signature: "msg_abc" },
{ type: "toolCall", id: "call_1", name: "read", arguments: {} },
{ type: "thinking", thinking: "hmm", thought_signature: "msg_xyz" },
];
const result = stripThoughtSignatures(input);
expect(result).toEqual([
{ type: "text", text: "hello" },
{ type: "toolCall", id: "call_1", name: "read", arguments: {} },
{ type: "thinking", thinking: "hmm" },
]);
});
it("handles empty array", () => {
expect(stripThoughtSignatures([])).toEqual([]);
});
it("handles null/undefined blocks in array", () => {
const input = [null, undefined, { type: "text", text: "hello" }];
const result = stripThoughtSignatures(input);
expect(result).toEqual([null, undefined, { type: "text", text: "hello" }]);
});
});
describe("sanitizeToolCallId", () => {
describe("strict mode (default)", () => {
it("keeps valid alphanumeric tool call IDs", () => {
expect(sanitizeToolCallId("callabc123")).toBe("callabc123");
});
it("strips underscores and hyphens", () => {
expect(sanitizeToolCallId("call_abc-123")).toBe("callabc123");
expect(sanitizeToolCallId("call_abc_def")).toBe("callabcdef");
});
it("strips invalid characters", () => {
expect(sanitizeToolCallId("call_abc|item:456")).toBe("callabcitem456");
});
});
describe("strict mode (alphanumeric only)", () => {
it("strips all non-alphanumeric characters", () => {
expect(sanitizeToolCallId("call_abc-123", "strict")).toBe("callabc123");
expect(sanitizeToolCallId("call_abc|item:456", "strict")).toBe("callabcitem456");
expect(sanitizeToolCallId("whatsapp_login_1768799841527_1", "strict")).toBe(
"whatsapplogin17687998415271",
);
});
});
describe("strict9 mode (Mistral tool call IDs)", () => {
it("returns alphanumeric IDs with length 9", () => {
const out = sanitizeToolCallId("call_abc|item:456", "strict9");
expect(out).toMatch(/^[a-zA-Z0-9]{9}$/);
});
});
it.each([
{
modeLabel: "default",
run: () => sanitizeToolCallId(""),
assert: (value: string) => expect(value).toBe("defaulttoolid"),
},
{
modeLabel: "strict",
run: () => sanitizeToolCallId("", "strict"),
assert: (value: string) => expect(value).toBe("defaulttoolid"),
},
{
modeLabel: "strict9",
run: () => sanitizeToolCallId("", "strict9"),
assert: (value: string) => expect(value).toMatch(/^[a-zA-Z0-9]{9}$/),
},
])("returns default for empty IDs in $modeLabel mode", ({ run, assert }) => {
assert(run());
});
});
describe("downgradeOpenAIReasoningBlocks", () => {
it("keeps reasoning signatures when followed by content", () => {
const input = [
{
role: "assistant",
content: [
{
type: "thinking",
thinking: "internal reasoning",
thinkingSignature: JSON.stringify({ id: "rs_123", type: "reasoning" }),
},
{ type: "text", text: "answer" },
],
},
];
// oxlint-disable-next-line typescript/no-explicit-any
expect(downgradeOpenAIReasoningBlocks(input as any)).toEqual(input);
});
it("drops orphaned reasoning blocks without following content", () => {
const input = [
{
role: "assistant",
content: [
{
type: "thinking",
thinkingSignature: JSON.stringify({ id: "rs_abc", type: "reasoning" }),
},
],
},
{ role: "user", content: "next" },
];
// oxlint-disable-next-line typescript/no-explicit-any
expect(downgradeOpenAIReasoningBlocks(input as any)).toEqual([
{ role: "user", content: "next" },
]);
});
it("drops object-form orphaned signatures", () => {
const input = [
{
role: "assistant",
content: [
{
type: "thinking",
thinkingSignature: { id: "rs_obj", type: "reasoning" },
},
],
},
];
// oxlint-disable-next-line typescript/no-explicit-any
expect(downgradeOpenAIReasoningBlocks(input as any)).toEqual([]);
});
it("keeps non-reasoning thinking signatures", () => {
const input = [
{
role: "assistant",
content: [
{
type: "thinking",
thinking: "t",
thinkingSignature: "reasoning_content",
},
],
},
];
// oxlint-disable-next-line typescript/no-explicit-any
expect(downgradeOpenAIReasoningBlocks(input as any)).toEqual(input);
});
it("is idempotent for orphaned reasoning cleanup", () => {
const input = [
{
role: "assistant",
content: [
{
type: "thinking",
thinkingSignature: JSON.stringify({ id: "rs_orphan", type: "reasoning" }),
},
],
},
{ role: "user", content: "next" },
];
// oxlint-disable-next-line typescript/no-explicit-any
const once = downgradeOpenAIReasoningBlocks(input as any);
// oxlint-disable-next-line typescript/no-explicit-any
const twice = downgradeOpenAIReasoningBlocks(once as any);
expect(twice).toEqual(once);
});
});
describe("normalizeTextForComparison", () => {
it("lowercases text", () => {
expect(normalizeTextForComparison("Hello World")).toBe("hello world");
});
it("trims whitespace", () => {
expect(normalizeTextForComparison(" hello ")).toBe("hello");
});
it("collapses multiple spaces", () => {
expect(normalizeTextForComparison("hello world")).toBe("hello world");
});
it("strips emoji", () => {
expect(normalizeTextForComparison("Hello 👋 World 🌍")).toBe("hello world");
});
it("handles mixed normalization", () => {
expect(normalizeTextForComparison(" Hello 👋 WORLD 🌍 ")).toBe("hello world");
});
});
describe("isMessagingToolDuplicate", () => {
it("returns false for empty sentTexts", () => {
expect(isMessagingToolDuplicate("hello world", [])).toBe(false);
});
it("returns false for short texts", () => {
expect(isMessagingToolDuplicate("short", ["short"])).toBe(false);
});
it("detects exact duplicates", () => {
expect(
isMessagingToolDuplicate("Hello, this is a test message!", [
"Hello, this is a test message!",
]),
).toBe(true);
});
it("detects duplicates with different casing", () => {
expect(
isMessagingToolDuplicate("HELLO, THIS IS A TEST MESSAGE!", [
"hello, this is a test message!",
]),
).toBe(true);
});
it("detects duplicates with emoji variations", () => {
expect(
isMessagingToolDuplicate("Hello! 👋 This is a test message!", [
"Hello! This is a test message!",
]),
).toBe(true);
});
it("detects substring duplicates (LLM elaboration)", () => {
expect(
isMessagingToolDuplicate('I sent the message: "Hello, this is a test message!"', [
"Hello, this is a test message!",
]),
).toBe(true);
});
it("detects when sent text contains block reply (reverse substring)", () => {
expect(
isMessagingToolDuplicate("Hello, this is a test message!", [
'I sent the message: "Hello, this is a test message!"',
]),
).toBe(true);
});
it("returns false for non-matching texts", () => {
expect(
isMessagingToolDuplicate("This is completely different content.", [
"Hello, this is a test message!",
]),
).toBe(false);
});
});