The Agent Skills Directory

[PROMPT_INJECTION] (LOW): Detected potential for Indirect Prompt Injection (Category 8) due to the skill's core function of analyzing application logs.
Ingestion points: The skill reads log/development.log using tail and grep to diagnose errors.
Boundary markers: Absent. There are no instructions to ignore or delimit instructions found within the logs.
Capability inventory: The skill is granted Bash, Read, Grep, and Glob tools, providing a significant surface for command execution or file modification if an injection is successful.
Sanitization: Absent. Log content is processed directly for analysis.
Context: An attacker who can influence log output (e.g., via web request parameters that get logged) could attempt to inject instructions for the agent to execute while it is debugging the logs.
[EXTERNAL_DOWNLOADS] (SAFE): The skill includes bundle install, which downloads external Ruby gems. While this involves network activity and package installation, it is the standard and expected behavior for a Rails development tool and is considered low risk in this specific context.
[COMMAND_EXECUTION] (SAFE): The skill uses Bash to run diagnostic commands like git log, bin/rails routes, and bundle check. These are restricted to standard development workflows and do not show signs of privilege escalation or malicious intent.

rails-debugger