The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes shell commands for environment detection, such as identifying if a repository is a fork or an original project by checking git remote configurations and parsing JSON manifests. It also grants specialized sub-agents access to the Bash tool to perform automated searching, syncing, and release management tasks within the local development environment.
[INDIRECT_PROMPT_INJECTION]: The skill processes external mod code and game source files, creating a potential attack surface where malicious instructions embedded in those files could influence agent behavior. 1. Ingestion points: Research agents read data from external directories including the game source and the local Mods folder as described in references/sub-agents.md. 2. Boundary markers: The skill does not explicitly define delimiters or specific 'ignore instructions' warnings for the content read from these external files. 3. Capability inventory: The code-writer and script-runner agents possess permissions for filesystem modification and command execution via the Bash, Write, and Edit tools (references/sub-agents.md). 4. Sanitization: No explicit sanitization or filtering of the ingested code content is defined before processing. This risk is inherent to code development tools and is managed by the skill's structured multi-phase reasoning workflow and isolated agent search boundaries.

balatro-mod-dev