The Agent Skills Directory

[SAFE]: The skill implements a robust routing logic between three specialized agents (PLG Activation Strategist, Interactive Demo Strategist, and Customer Marketing Lead) with clearly defined boundaries and handoff protocols.
[SAFE]: Includes explicit guardrails against hallucination, instructing the agents to use [NEEDS INPUT: …] markers rather than inventing metrics, customer names, or funding details.
[SAFE]: External resources and references used (Stripe documentation, FTC guidelines, Gartner research, and Navattic reports) are from well-known and reputable sources and are used neutrally for context rather than code execution.
[INDIRECT_PROMPT_INJECTION]: The skill ingests a user-controlled brand-context.md file and instructs agents that its contents may override defaults.
Ingestion points: Reads brand-context.md from the project root or .claude/ directory (SKILL.md).
Boundary markers: Explicitly instructs the agent to treat the file as binding and hand it to specialists (SKILL.md).
Capability inventory: The agents primarily perform text generation and routing; no dangerous shell execution or network operations are present in the provided scripts.
Sanitization: The skill relies on structured routing and explicit instructions to avoid inventing data, though it does not implement formal technical sanitization of the markdown file contents.

growth-ops