The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes multiple shell commands using the aws CLI and jq to interact with AWS CloudWatch and Logs. These include listing alarms, fetching metrics, and filtering log events. This is consistent with the skill's stated purpose.
[CREDENTIALS_UNSAFE]: The skill requires AWS_ACCESS_KEY_ID and AWS_SECRET_ACCESS_KEY to be present in the environment. While these are sensitive credentials, their use is standard for the intended functionality and no hardcoded secrets were detected.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection (Category 8) because it processes untrusted data as arguments for shell commands.
Ingestion points: User-provided inputs are used for parameters like --dimensions, --log-group-name, and --filter-pattern (e.g., in SKILL.md).
Boundary markers: No explicit delimiters or instructions to ignore embedded commands are present in the command templates.
Capability inventory: The skill possesses the capability to execute arbitrary subprocesses via the shell (aws CLI).
Sanitization: There is no evidence of sanitization or validation of inputs before they are passed to the shell, which could allow for command or argument injection if an attacker provides malicious log group names or patterns.

aws-cloudwatch