The Agent Skills Directory

[COMMAND_EXECUTION]: The skill's operating rules instruct the agent to offer a specific shell command (npx skills add wondelai/skills/<slug> --global) to the user if the required phase-specific skills are missing. These resources are hosted within the author's own namespace ('wondelai').
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it is designed to read and process external data stored in project documentation files (docs/CUSTOMER.md, docs/CREATE-BUSINESS-PLAN.md). Specifically, the agent is tasked with 'scoring transcripts' and reading 'every artifact' before proceeding, which could include malicious instructions embedded in interview notes.
Ingestion points: The agent reads from docs/CUSTOMER.md (Interview Evidence and Transcript scoring) and docs/CREATE-BUSINESS-PLAN.md at the start of every session.
Boundary markers: There are no explicit instructions to use delimiters or ignore embedded instructions when reading from the docs/ artifacts.
Capability inventory: The agent has the capability to write to the file system (docs/ directory) and invoke additional skills based on the contents of the tracker.
Sanitization: No sanitization or validation logic is defined for processing the content of customer transcripts or user-provided notes.

create-business