The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to process untrusted external data to drive autonomous decision-making and project execution cycles.
Ingestion points: The agent is instructed to investigate emerging approaches, profile accessible data, and read primary sources or evidence for discovery as defined in SKILL.md and references/data-quality.md.
Boundary markers: There are no explicit instructions to use delimiters or sanitization logic when processing external evidence to prevent the agent from following instructions embedded within that data.
Capability inventory: The skill manages high-privilege operations including software implementation (s-dev), quality assurance (s-qa), and repository modifications (git push).
Sanitization: The instructions lack specific requirements for sanitizing or escaping external content before it influences the agent's reasoning or execution path.
[COMMAND_EXECUTION]: The skill manages technical operations through Git, facilitating local and remote repository changes.
Evidence: references/git-delivery.md provides guidance for executing commit, push, and pull request operations to fulfill delivery requirements.
[DATA_EXFILTRATION]: The skill provides the agent with authorization to publish code and artifacts to remote repositories (references/git-delivery.md). While governed by authority checks, this mechanism creates a path for project data exfiltration if the agent's intent is compromised by malicious input.
[EXTERNAL_DOWNLOADS]: The skill references professional framework documentation from DAMA International (damadmbok.org) to support its data quality assessment capabilities. These references target established, well-known industry resources.

sflo-waydriver