Skip to main content

Squillify development is paused indefinitely. Read why, and discover what this experiment achieved. Read the story

MicroSkills Reference

Secret Detector

security.detect_secrets

ProductionHeuristic (Class C)Deterministicbest-effortfastv1.0.0
No network

Overview

You pass text, code, or config; we scan it with a curated, deterministic, ReDoS-safe rule pack (PEM keys, AWS access-key IDs, GitHub/Slack tokens, JWTs, bearer tokens, basic-auth URLs, generic credential assignments) plus optional Shannon-entropy checks in strict mode. This is preflight risk reduction — not Gitleaks/TruffleHog parity. We never return the full secret, only a redacted preview. Do not treat a clean result as proof the text is secret-free.

Secrets leak into code, config, and agent output all the time — you need a deterministic preflight before publishing, committing, or transmitting content.

Reliability

BEST_EFFORT with deterministic detectors (versioned rule pack): identical input yields the same findings on this deployment. A clean result is risk reduction, not certification. Input validation (not retryable): missing text; text too long; invalid contentKind or policy; maxFindings out of range.

Context footprint

Replaces ad-hoc, unreliable LLM secret-spotting with stable rules. A successful call costs 2 execution credits.

When to use it

Before committing, publishing, summarizing, or transmitting text or code.

As a sole compliance or DLP control — coverage is bounded and best-effort.

Scope

Note
Scans text, code, and config for likely secrets using deterministic, ReDoS-safe pattern detectors plus an optional Shannon-entropy check.

Non-goals

  • Does not guarantee complete secret removal (best-effort)
  • Does not call external scanners or networks
  • Does not return full secret values
  • Not Gitleaks/TruffleHog parity — a curated 8-detector pack, not hundreds of provider rules
  • Does not verify whether a matched value is a live/active credential
  • Does not tune the rule pack to contentKind — the same detectors run for every kind
  • Known false-negative classes: unlisted providers (GCP/Azure/Stripe/Twilio/…), obfuscated/base64-wrapped or line-split secrets, and low-entropy custom tokens

Properties

IdempotentRetry safeStreaming supportedFile input supportedRequires external service

Capabilities

analysisvalidation

Cost & billing

Flat rate per invocation

MetercallsBillable onsuccessCost per call2 credits

File handling

Accepts file_refProduces file_refMax input1.9 MB

How we handle your data

Sensitive processing
What this means

This skill is designed to detect, mask, redact, validate, or otherwise process sensitive data.

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.
  • Targets sensitive data
  • Output contains sensitive findings
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

FieldTypeRequiredDescription
textstringYesText, code, or configuration to scan.
contentKindstringNoContent label; omit it for text.
policystringNoDetector policy; omit it for standard.
includeLowConfidencebooleanNoWhether findings below 0.7 are included; omit it for false.
maxFindingsintegerNoFinding cap; omit it for 1,000.

Output schema

FieldTypeRequiredDescription
hasSecretsbooleanYesWhether findings survived filtering.
findingsarrayYesFrom 0 through 10,000 detector findings after filtering, each a closed non-secret receipt.
findingCountintegerNoReturned finding count.
truncatedbooleanNoWhether the finding cap omitted matches.

API contract

POST /v1/execute
curl
curl -sS -X POST "$BASE/v1/execute" -H "Authorization: Bearer $KEY" -H 'Content-Type: application/json' -d '{"skillKey":"security.detect_secrets","input":{"text":"aws_key = AKIAIOSFODNN7EXAMPLE"}}'
Integrations
TypeScript SDK
import { Squillify } from "@squillify/sdk";

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

const { output } = await sq.execute("security.detect_secrets", {"text":"aws_key = AKIAIOSFODNN7EXAMPLE"});

Errors

CodeHTTPRetryableDescription
VALIDATION_ERROR400NoThe request payload failed schema validation.
UNAUTHORIZED401NoMissing or invalid API key.
SKILL_NOT_ACCESSIBLE403NoThe API key is not granted access to this skill.
SKILL_NOT_AVAILABLE_FOR_PLAN403NoThis skill is not included in your current plan.
PLAN_ELIGIBILITY_UNAVAILABLE503YesPlan eligibility could not be decided. Retry after backoff.
SKILL_NOT_FOUND404NoThe requested skill or version does not exist.
TENANT_RATE_LIMITED429YesRequest rate limit exceeded. Retry after backoff.
QUOTA_EXCEEDED429YesPlan quota exhausted. Retryable once the quota period resets.
EXECUTION_TIMEOUT504YesExecution exceeded the allowed time budget.

Other limits

LimitValue
Recommended inline bytes32 KiB

Limitations

  • Does not guarantee complete secret removal (best-effort)
  • Does not call external scanners or networks
  • Does not return full secret values
  • Not Gitleaks/TruffleHog parity — a curated 8-detector pack, not hundreds of provider rules
  • Does not verify whether a matched value is a live/active credential
  • Does not tune the rule pack to contentKind — the same detectors run for every kind
  • Known false-negative classes: unlisted providers (GCP/Azure/Stripe/Twilio/…), obfuscated/base64-wrapped or line-split secrets, and low-entropy custom tokens
  • Guaranteed compliance-grade DLP (this is best-effort)
  • Detecting semantic PII like names/addresses (use text.mask_pii)
  • Validating that a specific known secret is present

How to call

POST /v1/execute (requires authentication)

Request bodyjson
{
  "skillKey": "security.detect_secrets",
  "input": {
    "text": "aws_key = AKIAIOSFODNN7EXAMPLE"
  }
}

Examples

AWS access key id detected

Inputjson
{
  "text": "aws_key = AKIAIOSFODNN7EXAMPLE"
}
Expected outputjson
{
  "findingCount": 1,
  "truncated": false,
  "hasSecrets": true,
  "findings": [
    {
      "confidence": 0.9,
      "kind": "aws_access_key_id",
      "column": 11,
      "detector": "aws_access_key_id",
      "startChar": 10,
      "line": 1,
      "redactedPreview": "AKIA…MPLE",
      "endChar": 30,
      "recommendation": "Rotate the AWS credential and remove it from source."
    }
  ]
}

Clean prose with no matches

Inputjson
{
  "text": "The quick brown fox."
}
Expected outputjson
{
  "findingCount": 0,
  "truncated": false,
  "hasSecrets": false,
  "findings": []
}
  • Preflight content before committing, publishing, or transmitting
  • Flag leaked credentials in agent-generated output
  • Gate documents before summarization or sharing

Avoid for

  • Guaranteed compliance-grade DLP (this is best-effort)
  • Detecting semantic PII like names/addresses (use text.mask_pii)
  • Validating that a specific known secret is present

Composable with

Safety & privacy

  • Runs without calling any external service.
  • Your payloads are never stored permanently.
  • Telemetry captures metadata only — never your payload contents.
  • We never train models on your payloads.
  • No human reviews your payloads.

Tags

verificationsecurityanalysis