The Agent Skills Directory

[SAFE]: The skill provides local tools for inspecting Aseprite (.ase/.aseprite) binary files. It implements a chunk-driven parser that follows the file specification, safely skipping unknown data sections.
[DYNAMIC_EXECUTION]: The parser uses the Node.js node:zlib module to handle compressed pixel data. The implementation in scripts/_lib/asset-tools.mjs includes a decompressLimited function that enforces a configurable safety limit (--max-decompress-mib, default 64MiB) to prevent decompression bomb attacks from maliciously crafted files.
[DATA_EXPOSURE]: The skill does not access sensitive system directories, credentials, or environment variables. It reads only the specific files provided as CLI arguments.
[COMMAND_EXECUTION]: The provided scripts are self-contained Node.js applications that do not spawn sub-shells or execute external commands derived from untrusted input.
[INDIRECT_PROMPT_INJECTION]: Although the skill processes external binary files, it acts as a passive parser and does not execute instructions found within the data. The binary reader uses explicit length checks to prevent buffer underruns during processing.

aseprite