Back to skill

Security audit

deepwiki

Security checks for vulnerabilities and agentic risk

Overview

This skill is a straightforward DeepWiki query helper that sends public repository targets and user questions to a documented third-party API, with privacy and prompt-injection cautions users should understand.

Install only if you are comfortable sending repository targets and questions to SkillBoss/HeyBossAI. Use it for public repositories as documented, avoid putting secrets or proprietary details in questions, and treat AI answers from untrusted repository documentation as advisory rather than authoritative.

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 (1)

T09 · Insecure Skill Coding Practices

Warning
Location
scripts/deepwiki.js:76
Finding

Untrusted Repository Documentation Is Embedded in a System Prompt

Content
View full analysis
Remediation
View remediation
\n${docText}\n\n\n` + `Question: ${extra}` } ] ``` This separation does not eliminate all model-level prompt-injection risk, but it creates a clearer trust boundary and materially reduces the likelihood that repository text will be treated as authoritative application policy. ]]>
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (6)

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
88% confidence
Finding

The skill requires an environment secret (SKILLBOSS_API_KEY) and invokes code, but the manifest does not declare an explicit tool scope such as permissions or allowed-tools. That omission weakens least-privilege controls and makes it harder for users or platforms to understand what capabilities the skill needs before execution.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The documentation instructs users to send repository identifiers and free-form questions to external services and mentions web scraping, but it does not clearly warn that prompts and queried content will leave the local system. This creates a real risk of unintended disclosure if a user includes private data, tokens, or sensitive repository details in questions.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
93% confidence
Finding

The skill explicitly transmits data to an external API endpoint and relies on a third-party web source for content retrieval. External transmission is expected for this skill's purpose, but it still represents a genuine security concern because user-supplied questions and metadata may be exposed to third parties or logged outside the local trust boundary.

Content

Scanner excerpt · SKILL.md (reported line 53)May include surrounding context.

text

## Notes
- Powered by SkillBoss API Hub (`https://api.heybossai.com/v1/pilot`).
- Web scraping fetches content from `https://deepwiki.com/<owner>/<repo>`.
- Works for public repositories only.
- The `ask` command uses SkillBoss API Hub's web scraping + LLM chat for AI-powered answers.

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% 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 · scripts/deepwiki.js (reported line 5)May include surrounding context.

js
const https = require('https');

const SKILLBOSS_API_KEY = process.env.SKILLBOSS_API_KEY;
const API_BASE = 'https://api.heybossai.com/v1';

const args = process.argv.slice(2);
const command = args[0];

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The ask command sends scraped repository documentation and the user's question to api.heybossai.com for LLM processing, but the script provides no explicit disclosure, consent prompt, or data-handling warning to the user. This creates a privacy and data-governance risk because repository content and prompts may contain sensitive information and are transmitted to a third party outside the local environment.

Content

No source excerpt is available for this finding.

Missing User Warnings

Low
Category
Not specified by scanner
Confidence
88% confidence
Finding

The structure and contents commands transmit repository-related DeepWiki URLs to a third-party scraping endpoint without clearly informing the user at execution time. While the data sent is typically less sensitive than full prompt/context content, it still exposes user activity and target repository information to an external service without transparent disclosure.

Content

No source excerpt is available for this finding.