The Agent Skills Directory

[DYNAMIC_EXECUTION]: The script scripts/office/soffice.py dynamically generates a C source file and compiles it into a shared object library (lo_socket_shim.so) using gcc at runtime.
[DYNAMIC_EXECUTION]: The skill uses the LD_PRELOAD environment variable to inject the compiled shim library into the soffice process to intercept and modify socket-related system calls, ensuring functionality in restricted environments.
[COMMAND_EXECUTION]: Multiple scripts (scripts/thumbnail.py, scripts/office/soffice.py, scripts/office/validators/redlining.py) use subprocess.run to execute external binaries including soffice, pdftoppm, git, and gcc.
[METADATA_POISONING]: The LICENSE.txt file contains a copyright notice for 'Anthropic, PBC', which is inconsistent with the provided author context ('Nagi-Inaba') and potentially misleading to users regarding the skill's origin.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted .pptx files, which creates a surface for indirect prompt injection if malicious instructions are embedded within slide text or metadata.
Ingestion points: scripts/office/unpack.py and scripts/thumbnail.py (via PPTX parsing).
Boundary markers: Absent; the instructions do not specify delimiters or warn the agent to ignore instructions found within slide content.
Capability inventory: Use of subprocess.run to execute system tools (soffice, pdftoppm, git).
Sanitization: The skill correctly utilizes defusedxml for XML parsing to mitigate XML External Entity (XXE) vulnerabilities.

pptx