Histórico de Auditoria
building-agents - 5 auditorias
Versão da auditoria 5
Mais recente Baixo RiscoJan 16, 2026, 07:42 PM
Pure prompt-based skill for Claude Code agent creation. Static findings are false positives: hex color codes misidentified as MD5 hashes, markdown code blocks misidentified as shell backticks, and CLI command examples misidentified as reconnaissance. The skill includes a Python validation script that safely reads user-specified files without network access or credential handling.
Problemas de Baixo Risco (1)
Fatores de risco
⚡ Contém scripts (1)
📁 Acesso ao sistema de arquivos (1)
⚙️ Comandos externos (1)
Versão da auditoria 4
Baixo RiscoJan 16, 2026, 07:42 PM
Pure prompt-based skill for Claude Code agent creation. Static findings are false positives: hex color codes misidentified as MD5 hashes, markdown code blocks misidentified as shell backticks, and CLI command examples misidentified as reconnaissance. The skill includes a Python validation script that safely reads user-specified files without network access or credential handling.
Problemas de Baixo Risco (1)
Fatores de risco
⚡ Contém scripts (1)
📁 Acesso ao sistema de arquivos (1)
⚙️ Comandos externos (1)
Versão da auditoria 3
Baixo RiscoJan 10, 2026, 11:46 AM
Pure prompt-based skill with a Python validation script. No network access, no credential handling, no file system overreach. The script only reads user-specified agent files for validation. All capabilities align with the stated purpose of agent creation and validation.
Problemas de Baixo Risco (1)
Fatores de risco
⚡ Contém scripts (1)
📁 Acesso ao sistema de arquivos (2)
⚙️ Comandos externos (1)
Versão da auditoria 2
Baixo RiscoJan 10, 2026, 11:46 AM
Pure prompt-based skill with a Python validation script. No network access, no credential handling, no file system overreach. The script only reads user-specified agent files for validation. All capabilities align with the stated purpose of agent creation and validation.
Problemas de Baixo Risco (1)
Fatores de risco
⚡ Contém scripts (1)
📁 Acesso ao sistema de arquivos (2)
⚙️ Comandos externos (1)
Versão da auditoria 1
Baixo RiscoJan 10, 2026, 11:46 AM
Pure prompt-based skill with a Python validation script. No network access, no credential handling, no file system overreach. The script only reads user-specified agent files for validation. All capabilities align with the stated purpose of agent creation and validation.