The Agent Skills Directory

[PROMPT_INJECTION]: No evidence of instructions attempting to bypass safety filters, override agent behavior, or extract system prompts. The instructions focus strictly on embedded engineering tasks.
[DATA_EXFILTRATION]: No sensitive file paths, credential harvesting, or unauthorized network operations were detected. The only external reference is the author's GitHub profile.
[OBFUSCATION]: The content is written in clear, human-readable text and standard C code. No Base64, zero-width characters, or homoglyph attacks are present.
[REMOTE_CODE_EXECUTION]: The skill does not perform remote script downloads or execution (e.g., curl | bash). It references standard static analysis and build tools (cppcheck, arm-none-eabi-size) used in local development workflows.
[COMMAND_EXECUTION]: No dangerous system commands are invoked. The workflow suggests standard developer tools for validation and optimization.
[INDIRECT_PROMPT_INJECTION]: While the skill assists in implementing code based on user specifications, it does not ingest untrusted data from external APIs or files in a way that would expose the agent to injection attacks.
[DYNAMIC_CONTEXT_INJECTION]: No use of the !command syntax was found in the SKILL.md file.
[PERSISTENCE]: No attempts to modify shell profiles, system services, or scheduled tasks were identified.

embedded-systems