The Agent Skills Directory

[Indirect Prompt Injection] (HIGH): The skill's primary function is to ingest and summarize longform untrusted content from external sources such as podcasts, videos, and articles.
Ingestion points: User-provided podcast transcripts, research papers, and web articles as mentioned in README.md.
Boundary markers: None identified. There are no instructions to the agent to ignore instructions embedded within the processed content.
Capability inventory: As a 'Claude Code' skill, the execution environment likely includes file system access and terminal command execution capabilities.
Sanitization: Absent. The skill provides formatting guidelines but no security filtering for external content.
[Unverifiable Dependencies] (MEDIUM): The README.md instructs users to clone a repository from https://github.com/WomenDefiningAI/claude-code-skills.git. This repository is not within the defined list of trusted organizations or repositories, posing a risk of supply chain attack if the remote content is modified.
[Metadata Poisoning] (LOW): The skill mentions version '1.7' and 'Last updated: 2025-Nov-16', which is a future date relative to current training data, though this is likely a typographical error rather than a malicious trigger.

shownotes-generator