The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill includes a Python-based data synchronization class that processes external data, which introduces a vulnerability surface where malicious instructions could be embedded in the processed content.
Ingestion points: The data parameter in the HybridDataSync.sync_to_cloud method within SKILL.md serves as a primary entry point for untrusted information.
Boundary markers: The code snippet lacks delimiters or instructional warnings to prevent the agent from interpreting embedded text as instructions.
Capability inventory: The skill logic includes network transmission (self.cloud_client.upload_batch) and persistent storage writes (self.local_store.store).
Sanitization: No input validation, filtering, or escaping is performed on the data before it is processed and stored.
[CREDENTIALS_UNSAFE]: The documentation for hybrid networking includes configuration examples with hardcoded placeholder secrets, which is a poor security practice that could lead to accidental exposure if copied verbatim.
Evidence: The aws_vpn_connection resource in SKILL.md contains hardcoded values for tunnel1_preshared_key = "secret1" and tunnel2_preshared_key = "secret2".

hybrid-cloud