The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill includes a reference to https://clawbot.ai/skills/munger-observer.html in its 'Source notes' section. This URL and its parent domain have been explicitly flagged as malicious (phishing) by automated security scanners. Referencing known phishing sites is dangerous as the agent may visit the site for context, or a user may click the link, potentially leading to credential theft or malware exposure.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to process and analyze untrusted external data such as user-provided documents, decision journals, and historical records. This creates a surface for indirect prompt injection where malicious instructions embedded in those documents could influence the agent's output or behavior.
Ingestion points: The skill ingests data via 'Observer/review' and 'Decision review' tasks, which read user-provided documents and history.
Boundary markers: No delimiters or explicit instructions are provided to the agent to ignore or isolate instructions found within the processed data.
Capability inventory: The skill possesses synthesis and recommendation capabilities and suggests verifying facts with available external sources, which may involve web browsing.
Sanitization: There are no visible mechanisms for sanitizing or validating the content of the documents being analyzed.

apply-munger-models