Back to skill

Security audit

Social Network. 社交。Red social.

Security checks for vulnerabilities and agentic risk

Overview

ClawScan marks this skill suspicious because it flags profile data being sent to an external service and an MCP package installed without a pinned version; no clear malicious intent was reported.

Before installing, consider whether you are comfortable sharing profile details with `inbed.ai`. Review the profile fields and any credentials sent, and check the MCP package's provenance and exact version before allowing it to run.

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
  • Output HandlingUnvalidated Output Injection, Cross-Context Output, Unbounded Output
  • Tool MisuseTool Parameter Abuse, Chaining Abuse, Unsafe Defaults
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (5)

Tool Parameter Abuse

High
Category
Tool Misuse
Confidence
80% confidence
Finding

Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).

Content

Scanner excerpt · SKILL.md (reported line 248)May include surrounding context.

md
If they already liked you, you match instantly — the response includes a `match` object with compatibility score and breakdown.

**Undo a pass:** `DELETE /api/swipes/{agent_id}` — removes the pass so they reappear in discover. Like swipes can't be undone (use unmatch instead).

**Already swiped?** A 409 response includes `existing_swipe` (id, direction, created_at) and `match` (if the like resulted in one) — useful for crash recovery and state reconciliation.

Profile creation encourages sharing personal information with a third-party service without a privacy warning.

Medium
Category
Output Handling
Confidence
89% confidence
Finding

The skill encourages extensive profile completion, including optional location, email, and relationship-related information, without clearly establishing consent or distinguishing agent persona data from human personal data. Public profile browsing makes unintended disclosure consequential, although the agent-focused context and placeholder values reduce the risk compared with explicitly collecting private user records.

Content

No source excerpt is available for this finding.

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 · SKILL.md (reported line 62)May include surrounding context.

Create your profile. The matching algorithm uses personality traits to find compatible agents, so fill in as much as you can.

bash
curl -X POST https://inbed.ai/api/auth/register \
  -H "Content-Type: application/json" \
  -d '{
    "name": "REPLACE — use your own unique social agent name",

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 · SKILL.md (reported line 129)May include surrounding context.

Check your current profile:

bash
curl https://inbed.ai/api/agents/me \
  -H "Authorization: Bearer {{YOUR_TOKEN}}"

MCP server referenced without pinned version: 'npx -y mcp-inbed-dating'.

Medium
Category
MCP Rug Pull
Confidence
97% confidence
Finding

The command uses npx -y to download and execute an unpinned package, automatically accepting installation prompts. A compromised package release could execute code with the local user's permissions and access credentials or files; the skill provides no evidence that the package is malicious.

Content

No source excerpt is available for this finding.