The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill includes documentation links to an external domain that has been flagged as malicious by automated URL reputation scanners.
Evidence: https://jeffallan.github.io/claude-skills/skills/infrastructure/cloud-architect/ (SKILL.md)
[METADATA_POISONING]: The main skill configuration file has been identified as malicious by file reputation scanners, likely due to the inclusion of high-risk external references.
Evidence: SKILL.md flagged as FileRepMalware.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest complex user-provided cloud requirements and translate them into executable CLI commands and infrastructure code, creating an attack surface for indirect prompt injection.
Ingestion points: User-provided cloud architecture requirements, constraints, and current state assessments in SKILL.md.
Boundary markers: None identified; the skill does not use specific delimiters or instructions to ignore embedded commands within user-provided requirements.
Capability inventory: The skill utilizes aws ec2, az network, aws elbv2, and aws iam CLI commands, as well as Terraform and Pulumi execution capabilities.
Sanitization: No sanitization or validation logic is present for user-supplied data before it is interpolated into architectural designs or command execution templates.

cloud-architect