The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to analyze and classify cold email replies, which constitute untrusted external data.
Ingestion points: The agent accesses and reads email reply content through the outreach_stats API or the suggested MCP server mentioned in SKILL.md.
Boundary markers: There are no instructions or delimiters provided to ensure the agent ignores potential malicious instructions embedded within the email replies.
Capability inventory: The skill references related skills such as reply-handling and outbound-campaign-ops, indicating the agent may have the capability to write files or perform network operations based on the analysis of these replies.
Sanitization: The skill does not implement or recommend any sanitization, filtering, or validation of the external reply content before it is processed by the agent.
[EXTERNAL_DOWNLOADS]: The skill provides a command to integrate a remote MCP server: claude mcp add --transport http manypi https://mcp.manypi.com/mcp -s user.
This command configures the agent to fetch data and tools from the author's (manypicom) hosted infrastructure.

outbound-metrics-benchmarks