diff --git a/.jules/sentinel.md b/.jules/sentinel.md index 3a1d237..a416b73 100644 --- a/.jules/sentinel.md +++ b/.jules/sentinel.md @@ -7,3 +7,8 @@ **Vulnerability:** The `CoverLetterGenerator` used a standard Jinja2 environment (intended for HTML/XML or plain text) to render LaTeX templates. This allowed malicious user input (or AI hallucinations) containing LaTeX control characters (e.g., `\input{...}`) to be injected directly into the LaTeX source, leading to potential Local File Inclusion (LFI) or other exploits. **Learning:** Jinja2's default `autoescape` is context-aware based on file extensions, but usually only for HTML/XML. It does NOT automatically escape LaTeX special characters. Relying on manual filters (like `| latex_escape`) in templates is error-prone and brittle, as developers might forget to apply them to every variable. **Prevention:** Always use a dedicated Jinja2 environment for LaTeX generation that enforces auto-escaping via a `finalize` hook (e.g., `tex_env.finalize = latex_escape`). This ensures *all* variable output is sanitized by default, providing defense-in-depth even if the template author forgets explicit filters. + +## 2024-05-18 - 🛡️ Sentinel: [CRITICAL/HIGH] Fix Remote Code Execution in PDF compilation +**Vulnerability:** Subprocess `pdflatex` / `pandoc` executed without `-no-shell-escape` and missing timeout. +**Learning:** Malicious inputs in Cover Letters or templates can execute arbitrary shell code via LaTeX `\write18` or `\immediate\write18` capabilities or cause Denial of Service by infinitely compiling. +**Prevention:** All PDF compilation tools must invoke `-no-shell-escape` explicitly, and utilize process `timeout` via `process.communicate(timeout=X)` and catching `subprocess.TimeoutExpired`. diff --git a/cli/generators/cover_letter_generator.py b/cli/generators/cover_letter_generator.py index aaf0b61..8da7871 100644 --- a/cli/generators/cover_letter_generator.py +++ b/cli/generators/cover_letter_generator.py @@ -771,12 +771,18 @@ def _compile_pdf(self, output_path: Path, tex_content: str) -> bool: try: # Use Popen with explicit cleanup to avoid double-free issues process = subprocess.Popen( - ["pdflatex", "-interaction=nonstopmode", tex_path.name], + ["pdflatex", "-interaction=nonstopmode", "-no-shell-escape", tex_path.name], stdout=subprocess.PIPE, stderr=subprocess.PIPE, cwd=tex_path.parent, ) - stdout, stderr = process.communicate() + try: + stdout, stderr = process.communicate(timeout=30) + except subprocess.TimeoutExpired: + process.kill() + stdout, stderr = process.communicate() + return False + if process.returncode == 0 or output_path.exists(): pdf_created = True except (subprocess.CalledProcessError, FileNotFoundError): @@ -787,11 +793,24 @@ def _compile_pdf(self, output_path: Path, tex_content: str) -> bool: # Fallback to pandoc try: process = subprocess.Popen( - ["pandoc", str(tex_path), "-o", str(output_path), "--pdf-engine=xelatex"], + [ + "pandoc", + str(tex_path), + "-o", + str(output_path), + "--pdf-engine=xelatex", + "--pdf-engine-opt=-no-shell-escape", + ], stdout=subprocess.PIPE, stderr=subprocess.PIPE, ) - stdout, stderr = process.communicate() + try: + stdout, stderr = process.communicate(timeout=30) + except subprocess.TimeoutExpired: + process.kill() + stdout, stderr = process.communicate() + return False + if process.returncode == 0 or output_path.exists(): pdf_created = True except (subprocess.CalledProcessError, FileNotFoundError):