Back to skill

Security audit

Technical Documentation Translator

Security checks for vulnerabilities and agentic risk

Overview

The skill is a coherent documentation translator, but it under-scopes sensitive file and URL handling enough that users should review it before installing.

Review before installing. Do not use folder mode on repositories or directories that may contain `.env`, credentials, private keys, or hidden configuration files, and avoid URL mode for internal systems or URLs containing tokens. Expect translated files to be created in the workspace.

Vulnerability Patterns
  • Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
  • 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
Findings (2)

T09 · Insecure Skill Coding Practices

Warning
Location
SKILL.md:16
Finding

Unsafe Processing and Duplication of Sensitive Environment Files

Content
View full analysis
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Warning
Location
SKILL.md:16
Finding

Unrestricted Arbitrary URL Retrieval

Content
View full analysis
Remediation
View remediation
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • System Prompt LeakageDirect Leakage, Indirect Extraction, Tool-Based Exfiltration
  • Rogue AgentSelf-Modification, Session Persistence
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Findings (4)

Direct Prompt Extraction

High
Category
System Prompt Leakage
Confidence
85% confidence
Finding

Skill contains instructions that could directly expose system prompts, internal rules, or hidden instructions to users or external parties.

Content

Scanner excerpt · SKILL.md (reported line 20)May include surrounding context.

md
Language direction is automatically determined, or you can manually specify `{source}→{target}` (e.g., `zh→en`).

## Output Rules

All output is written to the workspace root directory:

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The skill explicitly states that it writes translated output to the workspace root directory, but it does not prominently warn users about file-creation side effects or possible overwrites/clutter. This can lead to unintended persistence of sensitive translated content or unexpected modification of the workspace during normal use.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The skill accepts URLs and 'fetches then translates' content without any explicit privacy or network warning. Users may unknowingly cause outbound requests that expose internal URLs, tokens in query strings, or confidential document locations to remote systems.

Content

No source excerpt is available for this finding.

Session Persistence

Medium
Category
Rogue Agent
Confidence
75% confidence
Finding

Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.

Content

Scanner excerpt · SKILL.md (reported line 40)May include surrounding context.

md
| JSON | `.json` | Translate values only; skip non-text fields like `version`/`keywords`/`config`/`scripts`/`dependencies`/`license`/`main` |
| YAML/TOML | `.yaml` `.yml` `.toml` | Translate string values only; preserve keys and structure. For front matter, translate `name`/`description`/`title` but keep field names. |
| HTML | `.html` `.htm` `.xhtml` | Translate text inside `<body>`/`<p>`/`<h1~6>`/`<li>`/`<td>`/`<th>`/`<a>`/`<title>`/`<meta description>`; preserve tag attributes. Do not touch `<code>`, `<pre>`, `<script>`, `<style>`. |
| XML/SVG | `.xml` `.svg` `.plist` | Translate element text content and comments; preserve tags, attributes, namespaces, CDATA. For SVG, translate `<text>` content but preserve font/Aria attributes. |
| RST | `.rst` | Translate body text; preserve heading adornments (`===` `---`), directives (e.g., `.. code-block::`), roles (e.g., `:ref:`), and reference links. |
| AsciiDoc | `.adoc` `.asciidoc` `.ad` | Translate body text; preserve heading markers, block delimiters (`----`), macros (e.g., `include::`), and attributes (e.g., `{name}`). |
| CSV/TSV | `.csv` `.tsv` | Detect header row, translate data cells; preserve delimiters and quotes. |