The Agent Skills Directory

[DYNAMIC_CONTEXT_INJECTION]: The skill uses the dynamic context injection syntax to perform environment checks when the skill is loaded. These commands verify whether required system tools and local scripts are available.
Evidence: Found !command -v traceroute ..., `!`command -v mtr ..., and !test -f scripts/traceroute/trace-network-path.sh ...`` in SKILL.md.
[COMMAND_EXECUTION]: The instructions direct the agent to execute system binaries and local shell scripts to perform network path analysis and latency diagnostics.
Evidence: Use of traceroute, mtr, and scripts located in scripts/traceroute/ such as trace-network-path.sh and compare-routes.sh.
[INDIRECT_PROMPT_INJECTION]: The skill processes output from external network diagnostic tools, which represents an ingestion point for potentially untrusted data from the network (e.g., hop hostnames).
Ingestion points: Output from traceroute and mtr commands.
Boundary markers: Not explicitly defined in the skill instructions.
Capability inventory: The skill can execute local shell scripts and system networking binaries.
Sanitization: The skill documentation mentions that targets are validated against a scope configuration file (.pentest/scope.json) via a PreToolUse hook.

traceroute