CVE-2026-47119: Overly Broad Llm Instruction Inclusion
Using a broad wildcard pattern (like "**/*.md") to load and parse context or instructions can lead to prompt injection or malicious instruction execution if an attacker can introduce arbitrary files into the target directory. Constrain loaded files to explicitly known, specific filenames.
Provally CuratedPublic repositoryMediumMedium confidenceVerifiedApache-2.0
Python
Pythongreprules fetch cve-2026-47119-overly-broad-llm-instruction-inclusion --engine opengrepDescription
Using a broad wildcard pattern (like "**/*.md") to load and parse context or instructions can lead to prompt injection or malicious instruction execution if an attacker can introduce arbitrary files into the target directory. Constrain loaded files to explicitly known, specific filenames.
Community feedback
0 signals from signed-in users.
- Useful
- 0
- False positive
- 0
- Metadata
- 0