Back to skill

Security audit

Wiro AI for Kotlin

Security checks for vulnerabilities and agentic risk

Overview

The scanners rated this skill benign with medium confidence: they found no AIG issues and judged the flagged API references purpose-aligned, while noting an optional unpinned install command; Skillspector coverage was partial.

The verdict is benign, not risk-free, and confidence is medium. Before installing, consider whether you trust the referenced API services and review or pin the optional installer and skill source; Skillspector’s partial analysis leaves some evidence incomplete.

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (3)

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · SKILL.md (reported line 34)May include surrounding context.

kotlin
// Shipped apps: your backend adds the Wiro credentials
val client = WiroClient(
    proxyUrl = "https://api.myapp.com/wiro/v1",
    headers = mapOf("Authorization" to "Bearer app-token"),
)

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · SKILL.md (reported line 45)May include surrounding context.

val client = WiroClient(apiKey = "YOUR_API_KEY", apiSecret = "YOUR_API_SECRET")

text

- Proxy mode stores no Wiro key. It sends the normal REST paths (`/Run/{owner}/{project}`, `/Task/Detail`, `/Task/Cancel`, `/Task/Kill`, `/File/Upload`, `/Tool/List`, `/Tool/Detail`, `/Tool/Explore`) to `proxyUrl` with your `headers`. Your backend checks its own user, adds the Wiro auth headers, forwards the call to `https://api.wiro.ai/v1` and returns Wiro's response unchanged.
- Pass `apiSecret` only for a project that uses signature auth (https://wiro.ai/docs/authentication).
- Name the first argument (`apiKey =` or `proxyUrl =`), since both constructors take a `String` first.
- Optional settings: `requestTimeout`, `pollInterval`, `retryPolicy`, `logger`, and `okHttpClient` to share your app's OkHttp client. The SDK's default client sets `retryOnConnectionFailure(false)`; set it on yours too, so OkHttp does not retry calls after connection failures on its own.

MCP server referenced without pinned version: 'npx skills'.

Medium
Category
MCP Rug Pull
Confidence
88% confidence
Finding

The suggested npx skills add command does not pin the skills package version or an immutable revision of the remote skill repository. Running it could execute a compromised future CLI release or install altered agent instructions, although this is an optional documentation reference rather than an automatically executed step. The static finding incorrectly describes this as an MCP server version issue; the actual concern is an unpinned installation workflow.

Content

No source excerpt is available for this finding.