MaxwellCalkin

Scan Tool Call — Development skill for Claude Code

Development community

/scan-tool-call — Check if a tool call is safe.

How to install Scan Tool Call

Installs to ~/.claude/skills/maxwellcalkin-sentinel-ai-scan-tool-call/SKILL.md

Terminal

mkdir -p ~/.claude/skills/maxwellcalkin-sentinel-ai-scan-tool-call && curl -fsSL https://raw.githubusercontent.com/MaxwellCalkin/sentinel-ai/HEAD/commands/scan-tool-call.md -o ~/.claude/skills/maxwellcalkin-sentinel-ai-scan-tool-call/SKILL.md

Restart Claude Code, or start a new session, for it to be picked up.

What Scan Tool Call does

/scan-tool-call — Check if a tool call is safe

Scan a tool call for dangerous operations before execution.

Usage

/scan-tool-call

Behavior

Use the Sentinel AI `scan_tool_call` MCP tool to:

Parse the tool name and JSON arguments
Check for dangerous shell commands (rm -rf, credential access, etc.)
Check for data exfiltration patterns (curl to external servers, etc.)
Check for privilege escalation attempts
Report risk level and whether the

Alternatives in Development

Om Tidy — Self-maintenance pass — acts on every hygiene flag: archives completed work, groups loose clusters, splits ove 4.6k ★
Claude-code-safety-net Rules Reference — Custom rules reference for defining safe/dangerous command patterns 1.2k ★
Claude Keysmith — Managed Claude Code instruction deployment with safe CLI recovery and an unsigned macOS/Windows desktop beta 696 ★