The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill directs the agent to research external sources for current pricing, integration status, and deliverability benchmarks for several third-party platforms. Processing content from these external sites introduces a surface for indirect prompt injection if the sites contain malicious instructions. 1. Ingestion points: Phase 2 (Research) instructs the agent to visit and process data from vendor and review websites. 2. Boundary markers: No specific delimiters or instructions to ignore embedded commands are provided for the research phase. 3. Capability inventory: The skill includes a local validation script (scripts/check-output.py) but does not include network-active scripts or operations that access sensitive local files. 4. Sanitization: No sanitization of researched content is specified before it is used to generate recommendations.
[COMMAND_EXECUTION]: The skill references and includes a local utility script, scripts/check-output.py, which is intended to be run by the agent to validate its own output. This script uses standard libraries to perform benign text analysis and does not represent a high-risk command execution vector.

sending-platforms