Skip to main content
Glama

DollhouseMCP

by DollhouseMCP
ISSUE_BACKTICK_FALSE_POSITIVE.md1.63 kB
# Issue: Security Validator Blocks Markdown Code Formatting ## Problem The content validator is blocking legitimate markdown content that uses backticks for code formatting. This prevents installation of valid content like the Roundtrip Test Skill. ## Root Cause In `src/security/contentValidator.ts` line 62: ```javascript { pattern: /`[^`]+`/g, severity: 'critical', description: 'Backtick command execution' }, ``` This pattern treats ALL backticks as potential command execution, even when they're just markdown code formatting. ## Example of False Positive The Roundtrip Test Skill contains legitimate markdown like: ```markdown Install: `install_content "library/skills/roundtrip-test-skill.md"` Details: `get_collection_content "library/skills/roundtrip-test-skill.md"` ``` This is blocked with error: "Critical security threat detected in persona content" ## Impact - Cannot install skills/templates that contain code examples - Cannot use backticks for inline code in any content - Blocks legitimate technical documentation ## Solution Needed The pattern needs to distinguish between: 1. **Shell command execution**: Real backticks in a shell context 2. **Markdown formatting**: Backticks used for code display Possible approaches: - Check context (is this in a YAML field vs markdown content?) - Look for actual shell patterns inside backticks - Allow single-line backticks but block multi-line - Whitelist markdown code blocks ## Temporary Workaround Remove the backtick pattern entirely (line 62) until a better solution is implemented. --- *Discovered: August 11, 2025* *Blocks: Roundtrip workflow testing*

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/DollhouseMCP/DollhouseMCP'

If you have feedback or need assistance with the MCP directory API, please join our Discord server