The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill possesses an indirect prompt injection attack surface.
Ingestion points: The 'Theme' and 'Material' input fields in SKILL.md explicitly accept user-provided content, including external URLs.
Boundary markers: No specific delimiters or instructions to ignore embedded commands within the processed materials are defined.
Capability inventory: The skill executes local scripts (skills/shared/scripts/wordcount.py) and writes content to the outputs/ directory.
Sanitization: There is no evidence of sanitization or validation of the content retrieved from external URLs before it is processed by the agent.
[COMMAND_EXECUTION]: The skill performs automated shell command execution for utility purposes.
Evidence: Step 6 in SKILL.md instructs the agent to execute python3 skills/shared/scripts/wordcount.py count to validate platform-specific character limits. This is a local execution of a shared utility script.
[SAFE]: The skill references several external GitHub repositories in EASEL-META.md (e.g., from users nashsu, maomao52088, op7418). These are documented neutrally as methodology references and do not involve remote code execution or unauthorized data access.

social-content