Back to skill

Security audit

Cospark

Security checks for vulnerabilities and agentic risk

Overview

The scanners rate this skill benign because its disclosed Cospark ad research and media tools align with its stated purpose, though the review is partial and notes a documented external service connection.

Consider whether you are comfortable connecting to Cospark and sending the content needed for its tools to that service. The verdict is benign, but not risk-free, and the scan is partial because a referenced local path was not found.

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
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (4)

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 19)May include surrounding context.

md
If Cospark tools are already available, skip setup. Otherwise add a remote Streamable HTTP MCP server:

- Name: `cospark`
- URL: `https://api.cospark.so/mcp`
- Authentication: browser OAuth with the user's Cospark account. No API key.
- Setup guide: https://www.cospark.so/mcp

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 23)May include surrounding context.

md
- Authentication: browser OAuth with the user's Cospark account. No API key.
- Setup guide: https://www.cospark.so/mcp

For Claude Code: `claude mcp add --transport http cospark https://api.cospark.so/mcp`, then use `/mcp` to sign in. For other clients, use their remote MCP configuration and OAuth login flow. Installing this skill alone does not connect the server. If your client cannot add tools in the current session, give the setup steps and resume in a session that has the connection.

For Hermes, add the connection using its native MCP setup: `hermes mcp add cospark --url https://api.cospark.so/mcp --auth oauth`. Complete browser sign-in and save the connection. Use `hermes mcp login cospark` to reconnect an existing configured server, then start a new session. Let Hermes own token storage and refresh. If login is required in an unattended run, report that setup is needed; do not loop or read another client's credentials.

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 25)May include surrounding context.

md
For Claude Code: `claude mcp add --transport http cospark https://api.cospark.so/mcp`, then use `/mcp` to sign in. For other clients, use their remote MCP configuration and OAuth login flow. Installing this skill alone does not connect the server. If your client cannot add tools in the current session, give the setup steps and resume in a session that has the connection.

For Hermes, add the connection using its native MCP setup: `hermes mcp add cospark --url https://api.cospark.so/mcp --auth oauth`. Complete browser sign-in and save the connection. Use `hermes mcp login cospark` to reconnect an existing configured server, then start a new session. Let Hermes own token storage and refresh. If login is required in an unattended run, report that setup is needed; do not loop or read another client's credentials.

## Load the current workflow

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 · agents/openai.yaml (reported line 11)May include surrounding context.

yaml
value: "cospark"
      description: "Cospark ad research, media generation, inspection, voiceover, and project tools"
      transport: "streamable_http"
      url: "https://api.cospark.so/mcp"