mcpscan good first issues (Python)

feat: scan MCP client config files with mcpServers 3 days ago

enhancement help wanted

0

Security scanner for MCP (Model Context Protocol) servers. Detects prompt injection, unsafe auth, path traversal and over-broad permissions. Zero dependencies.

Python

#agentic-ai#ai-security#anthropic#claude#developer-tools#llm#mcp#model-context-protocol#owasp#prompt-injection#red-team#scanner#security#tool-poisoning

refactor: build scan JSON output in one place 3 days ago

good first issue refactor

obielin/mcpscan

0

Security scanner for MCP (Model Context Protocol) servers. Detects prompt injection, unsafe auth, path traversal and over-broad permissions. Zero dependencies.

Python

#agentic-ai#ai-security#anthropic#claude#developer-tools#llm#mcp#model-context-protocol#owasp#prompt-injection#red-team#scanner#security#tool-poisoning

feat: detect hidden instructions in tool descriptions 3 days ago

enhancement help wanted

obielin/mcpscan

0

Security scanner for MCP (Model Context Protocol) servers. Detects prompt injection, unsafe auth, path traversal and over-broad permissions. Zero dependencies.

Python

#agentic-ai#ai-security#anthropic#claude#developer-tools#llm#mcp#model-context-protocol#owasp#prompt-injection#red-team#scanner#security#tool-poisoning

fix: MR-004 misses an open path field when another field has a pattern 3 days ago

bug good first issue

obielin/mcpscan

0

Security scanner for MCP (Model Context Protocol) servers. Detects prompt injection, unsafe auth, path traversal and over-broad permissions. Zero dependencies.

Python

#agentic-ai#ai-security#anthropic#claude#developer-tools#llm#mcp#model-context-protocol#owasp#prompt-injection#red-team#scanner#security#tool-poisoning

fix: malformed manifests crash the scan or skip every rule 3 days ago

bug help wanted

obielin/mcpscan

0

Security scanner for MCP (Model Context Protocol) servers. Detects prompt injection, unsafe auth, path traversal and over-broad permissions. Zero dependencies.

Python

#agentic-ai#ai-security#anthropic#claude#developer-tools#llm#mcp#model-context-protocol#owasp#prompt-injection#red-team#scanner#security#tool-poisoning

fix: a HIGH finding does not fail the scan 3 days ago

bug help wanted

obielin/mcpscan

0

Security scanner for MCP (Model Context Protocol) servers. Detects prompt injection, unsafe auth, path traversal and over-broad permissions. Zero dependencies.

Python

#agentic-ai#ai-security#anthropic#claude#developer-tools#llm#mcp#model-context-protocol#owasp#prompt-injection#red-team#scanner#security#tool-poisoning

fix: directory scans count top-level manifest files twice 3 days ago

bug good first issue

obielin/mcpscan

0

Security scanner for MCP (Model Context Protocol) servers. Detects prompt injection, unsafe auth, path traversal and over-broad permissions. Zero dependencies.

Python

#agentic-ai#ai-security#anthropic#claude#developer-tools#llm#mcp#model-context-protocol#owasp#prompt-injection#red-team#scanner#security#tool-poisoning

feat: scan a live MCP server over stdio 3 days ago

enhancement help wanted

obielin/mcpscan

0

Security scanner for MCP (Model Context Protocol) servers. Detects prompt injection, unsafe auth, path traversal and over-broad permissions. Zero dependencies.

Python

#agentic-ai#ai-security#anthropic#claude#developer-tools#llm#mcp#model-context-protocol#owasp#prompt-injection#red-team#scanner#security#tool-poisoning

feat: add SARIF output to mcpscan scan 3 days ago

enhancement help wanted

obielin/mcpscan

0

Security scanner for MCP (Model Context Protocol) servers. Detects prompt injection, unsafe auth, path traversal and over-broad permissions. Zero dependencies.

Python

#agentic-ai#ai-security#anthropic#claude#developer-tools#llm#mcp#model-context-protocol#owasp#prompt-injection#red-team#scanner#security#tool-poisoning

fix: scan_file fails on JSONL tool listings 3 days ago

bug good first issue

obielin/mcpscan

0

Security scanner for MCP (Model Context Protocol) servers. Detects prompt injection, unsafe auth, path traversal and over-broad permissions. Zero dependencies.

Python

#agentic-ai#ai-security#anthropic#claude#developer-tools#llm#mcp#model-context-protocol#owasp#prompt-injection#red-team#scanner#security#tool-poisoning

obielin/mcpscan

Open Issues Need Help