The Agent Skills Directory

[PROMPT_INJECTION]: No prompt injection patterns or attempts to override agent safety guidelines were detected in the instructions.
[DATA_EXFILTRATION]: The skill does not perform any network operations or access sensitive system files like credentials or SSH keys. All technical references point to well-known and trusted educational sites such as kernel.org and osdev.org.
[REMOTE_CODE_EXECUTION]: There are no patterns involving the download and execution of remote scripts. The skill uses local Python and C files for demonstration purposes.
[INDIRECT_PROMPT_INJECTION]: While the skill processes binary 'blobs' representing PCIe configuration space, the provided scripts include robust security practices such as cycle detection and bounds checking to prevent resource exhaustion or crashes from malformed input data.
[DYNAMIC_EXECUTION]: The skill includes C source code and instructions for compilation using GCC. This is standard behavior for teaching system programming and does not involve executing untrusted or dynamically generated code from external sources.
[COMMAND_EXECUTION]: The skill documentation mentions standard hardware diagnostic tools like lspci and setpci, which are used locally for hardware verification and do not pose a security risk in this context.

pcie-config-space