The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The scripts/install.sh script performs a download and direct execution of a shell script from https://plannotator.ai/install.sh by piping the output of curl into sh. This is a significant security risk as the remote content is not verified before execution and the source domain is not a recognized trusted vendor.
[DATA_EXFILTRATION]: In scripts/phase-gate-review.sh, the skill captures the full text content of documents located in the docs/ folder, such as Product Requirements Documents (PRDs) and architecture specifications, and transmits them to an external service via the plannotator submit command. This represents a potential exposure of sensitive project intellectual property.
[EXTERNAL_DOWNLOADS]: The installation process involves fetching and running software from plannotator.ai, which is not classified as a trusted organization or well-known technology service in this context.
[COMMAND_EXECUTION]: The skill relies on multiple shell scripts to automate its workflow. For example, scripts/phase-gate-review.sh executes commands using user-supplied file paths and document titles without sufficient validation, creating a potential vector for command injection.

bmad