Historique des audits

This is a legitimate academic research tool that makes API calls to OpenRouter for literature searches. All 146 static findings are false positives: (1) API key environment reads are standard authentication for OPENROUTER_API_KEY, (2) network calls target the legitimate OpenRouter API endpoint, (3) 'external commands' flagged in SKILL.md are documentation examples showing command syntax, not runtime code execution, and (4) 'weak cryptographic algorithm' mentions are ML architecture names (transformer, RNN) misidentified as cryptographic algorithms. No malicious patterns detected after intent analysis.

This is a legitimate academic research tool that makes API calls to OpenRouter for literature searches. All 146 static findings are false positives: (1) API key environment reads are standard authentication for OPENROUTER_API_KEY, (2) network calls target the legitimate OpenRouter API endpoint, (3) 'external commands' flagged in SKILL.md are documentation examples showing command syntax, not runtime code execution, and (4) 'weak cryptographic algorithm' mentions are ML architecture names (transformer, RNN) misidentified as cryptographic algorithms. No malicious patterns detected after intent analysis.

This is a legitimate academic research tool that makes API calls to OpenRouter for literature searches. The static scanner flagged 127 potential issues, but all are false positives: (1) API key environment reads are standard authentication, (2) network calls target the legitimate OpenRouter API, (3) 'external commands' flagged in SKILL.md are documentation examples, not code, and (4) 'weak cryptographic algorithm' mentions are ML architecture names (transformer, RNN) misidentified as crypto. No malicious patterns detected.

Legitimate research lookup skill. Makes API calls to OpenRouter for Perplexity Sonar models. Reads OPENROUTER_API_KEY for authentication. No suspicious patterns detected.