The Agent Skills Directory

[COMMAND_EXECUTION]: The script scripts/create_layout_issue_bundle.py executes system commands using subprocess.run with list-based arguments. This approach is used to invoke utilities like pdftoppm, pdftotext, and ImageMagick for document processing. The use of argument lists instead of raw shell strings prevents command injection vulnerabilities.\n- [SAFE]: No network exfiltration, credential harvesting, or obfuscation was detected. The skill's behavior matches its described purpose of local artifact creation. All external resources used (e.g., ImageMagick, Poppler) are standard system tools expected in the skill's operating environment.\n- [PROMPT_INJECTION]: The skill represents a surface for indirect prompt injection as it ingests content from external files (PDFs, LaTeX source, and logs) and formats them into a prompt.md intended for other agents. \n
Ingestion points: scripts/create_layout_issue_bundle.py reads from PDF files, .tex source files, and compile logs.\n
Boundary markers: The generated prompt.md uses Markdown headers and code blocks to separate user instructions from ingested snippets.\n
Capability inventory: The skill has the capability to write files to the local repository and execute local command-line tools.\n
Sanitization: Inputs used for file paths and identifiers are sanitized using a slugify function to prevent directory traversal.

latex-layout-issue-bundler