The Agent Skills Directory

[DYNAMIC_EXECUTION]: The module scripts/office/soffice.py performs dynamic code generation and execution to handle environment restrictions.
It contains a hardcoded C source string (_SHIM_SOURCE) designed to shim Unix socket system calls.
At runtime, it writes this source to a randomized temporary directory, compiles it into a shared object using gcc, and then injects it into the soffice process using the LD_PRELOAD environment variable.
This mechanism is a legitimate compatibility shim for sandboxed environments and includes security measures to prevent local file swapping attacks.
[COMMAND_EXECUTION]: The skill executes several external binaries via subprocess.run to perform its primary functions.
scripts/recalc.py and scripts/office/soffice.py execute soffice (LibreOffice) to recalculate workbook formulas and convert documents.
scripts/office/soffice.py executes gcc to compile the socket shim when environment restrictions are detected.
scripts/office/validators/redlining.py executes git diff to compare text content and identify untracked changes in Word documents.
In all cases, the scripts build minimal environments for these processes to prevent the accidental exposure of sensitive environment variables or credentials.
[INDIRECT_PROMPT_INJECTION]: The skill has a large attack surface as it processes untrusted Office documents (.xlsx, .docx, .pptx) which are parsed and then converted into Markdown or JSON for the agent.
Ingestion points: scripts/office/validate.py and scripts/recalc.py read external files into memory.
Sanitization: The skill proactively mitigates common XML vulnerabilities by using the defusedxml library for parsing all Office XML components.
Capability inventory: The skill has extensive file-writing and subprocess execution capabilities used for repair and validation tasks.

xlsx