The Agent Skills Directory

[SAFE]: The skill consists of instructional prompts designed to help a user create sales documentation. It does not contain any executable code, remote script downloads, or obfuscated patterns. Its behavior is consistent with its stated purpose.
[DATA_EXFILTRATION]: The skill reads from project-specific files (e.g., my-gtm-context.md and various files in the outputs/ directory) to provide contextual advice. No sensitive system paths or credentials are accessed, and there are no network operations to exfiltrate data.
[PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection as it processes data from external project files. However, given the context of a local multi-phase methodology, this is an intended feature for maintaining strategic consistency.
Ingestion points: Reads my-gtm-context.md and prior phase deliverables from the outputs/ folder (Phases 3, 5, 6, 7, 10, 11).
Boundary markers: Absent; instructions do not specify delimiters or warnings for the agent when reading these files.
Capability inventory: Limited to generating markdown text files in the outputs/ directory.
Sanitization: Absent; no explicit validation of the ingested content is performed.

executing-sales