The Agent Skills Directory

[PROMPT_INJECTION]: The skill is designed to ingest and analyze untrusted external data such as code, architecture proposals, and bug reports. This establishes a surface area for indirect prompt injection, where malicious instructions embedded in processed files could attempt to influence the agent's behavior. However, the risk is mitigated as the skill does not define any executable capabilities (such as network access, file system modifications, or subprocess execution) that could be exploited via such an injection.
[NO_CODE]: This skill consists entirely of natural language instructions and configuration metadata. It does not include any executable scripts, binary files, or dependencies, significantly reducing the potential for traditional malware or remote code execution vectors.

analysis-swarm