The Agent Skills Directory

[SAFE]: The skill consists entirely of markdown-based instructions for the agent to follow. It does not request any tool permissions (allowed-tools is absent) or include any executable code.
[PROMPT_INJECTION]: The skill has a surface for indirect prompt injection as it is designed to process untrusted data such as transcripts and video descriptions.
Ingestion points: Data is ingested through the [paste] placeholder in the SKILL.md instruction block.
Boundary markers: Absent; the skill does not use delimiters or clear markers to separate user instructions from external content.
Capability inventory: No tools or file system access are defined in the frontmatter, limiting the potential impact of an injection to text output.
Sanitization: Absent; the content is processed directly as text.

reel-analyzer