Check Bidi Safety

localization.bidi_safety_check

ProductionEngine wrapper (Class A)Deterministicstrictfastv1.0.0

No network

Overview

You send text and a context. We count explicit bidi controls, walk embeddings and isolates, and flag a leftover pop, a deprecated LRO or RLO, a control inside an identifier, or a Trojan-Source pattern in source. Arabic and Hebrew letters without controls stay safe. We do not strip, render or call it malware. A successful call costs 1 execution credit.

Trojan Source and unbalanced isolates change what a reviewer sees. A model cannot prove stack balance from a glance at the string.

Reliability

Same text, same context, same ICU version, same answer. Findings always include utf16Offset and codePointOffset. Structural problems return safe false with diagnostics rather than a typed failure.

Context footprint

One string in. A successful call costs 1 execution credit.

When to use it

When you hold a string and need to know whether its bidi controls are safe for that context.

When you want to generate pseudo-RTL copy, or to detect language.

Scope

Note

Statically flags unbalanced, deprecated or identifier-unsafe bidirectional controls in one string.

Non-goals

Does not render the visual order
Does not strip or rewrite controls
Does not issue a malware verdict

Properties

IdempotentRetry safeStreaming supportedFile input supportedRequires external service

Capabilities

analysisvalidation

Cost & billing

Flat rate per invocation

MetercallsBillable onsuccessCost per call1 credit

File handling

Accepts file_refProduces file_refMax input1 MB

How we handle your data

Sensitive input, safe output

What this means

This skill may receive sensitive content, but it does not target sensitive data and returns only a structured non-sensitive result.

Squillify guarantees

Runs inside Squillify
This skill runs inside Squillify.
Payloads are not sent to an external service by this skill.
Squillify does not permanently store submitted payloads.
Telemetry records metadata only, not submitted content.
Submitted payloads are not used for training.
Submitted payloads are not reviewed by humans.
This skill does not target sensitive data.
The response does not contain submitted content.

Your responsibility

Squillify does not determine whether submitted content is sensitive. Your application is responsible for deciding what data it is allowed to submit. Your application is responsible for securing the response.

Read the Microskill Data Handling Policy

Input schema

Field	Type	Required	Description
text	string	Yes	String to scan for explicit bidirectional controls and unbalanced embeddings.
context	string	Yes	PLAIN_TEXT, IDENTIFIER or SOURCE_CODE; each applies a different control policy.

Output schema

Field	Type	Required	Description
safe	boolean	Yes	True when every diagnostic is a warning or the list is empty; false when an error exists.
controlCounts	object	Yes	Closed counts of explicit bidirectional controls: lre rle pdf lro rlo lri rli fsi pdi lrm rlm alm.
diagnostics	array	Yes	From 0 through 500 diagnostics sorted by path, offset, severity and code; empty when nothing was flagged.
_meta	object	Yes	Closed ICU and CLDR versions used for this answer, with icuVersion and cldrVersion strings.

API contract

POST /v1/execute

curl

curl -sS -X POST "$BASE/v1/execute" -H "Authorization: Bearer $KEY" -H 'Content-Type: application/json' -d '{"skillKey":"localization.bidi_safety_check","input":{"text":"Hello","context":"PLAIN_TEXT"}}'

Integrations

TypeScript SDK

import { Squillify } from "@squillify/sdk";

const sq = new Squillify({ apiKey: process.env.SQUILLIFY_API_KEY });

const { output } = await sq.execute("localization.bidi_safety_check", {"context":"PLAIN_TEXT","text":"⁧Hello⁩"});

Errors

Code	HTTP	Retryable	Description
VALIDATION_ERROR	400	No	The request payload failed schema validation.
UNAUTHORIZED	401	No	Missing or invalid API key.
SKILL_NOT_ACCESSIBLE	403	No	The API key is not granted access to this skill.
SKILL_NOT_AVAILABLE_FOR_PLAN	403	No	This skill is not included in your current plan.
PLAN_ELIGIBILITY_UNAVAILABLE	503	Yes	Plan eligibility could not be decided. Retry after backoff.
SKILL_NOT_FOUND	404	No	The requested skill or version does not exist.
TENANT_RATE_LIMITED	429	Yes	Request rate limit exceeded. Retry after backoff.
QUOTA_EXCEEDED	429	Yes	Plan quota exhausted. Retryable once the quota period resets.
EXECUTION_TIMEOUT	504	Yes	Execution exceeded the allowed time budget.

Other limits

Limit	Value
Recommended inline bytes	32 KiB

Limitations

Does not render the visual order
Does not strip or rewrite controls
Does not issue a malware verdict
Turning English copy into a pseudo-RTL string
Detecting the language of a paragraph

How to call

POST /v1/execute (requires authentication)

Request bodyjson

{
  "skillKey": "localization.bidi_safety_check",
  "input": {
    "context": "PLAIN_TEXT",
    "text": "⁧Hello⁩"
  }
}

Examples

Balanced isolate in plain text

Inputjson

{
  "context": "PLAIN_TEXT",
  "text": "⁧Hello⁩"
}

Expected outputjson

{
  "diagnostics": [],
  "safe": true,
  "_meta": {
    "icuVersion": "76.1",
    "cldrVersion": "46.0"
  },
  "controlCounts": {
    "lre": 0,
    "rle": 0,
    "pdf": 0,
    "lro": 0,
    "rlo": 0,
    "lri": 0,
    "rli": 1,
    "fsi": 0,
    "pdi": 1,
    "lrm": 0,
    "rlm": 0,
    "alm": 0
  }
}

Tip

A matched RLI/PDI pair is safe in plain text

Right-to-left override in source