The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill relies on external data sources that could be manipulated to influence agent behavior through indirect prompt injection.
Ingestion points: The agent reads the organization's praxis profile from praxis-profile.md, AGENTS.md, or CLAUDE.md to resolve organizational outcomes, compute tiers, and expert roles (SKILL.md).
Boundary markers: The instructions lack explicit boundary markers or directives to treat data from the praxis profile as untrusted when evaluating gate questions or failure modes.
Capability inventory: The skill has the capability to write to the file system to create or update praxis-profile.md and to modify project instruction files like AGENTS.md or CLAUDE.md (SKILL.md).
Sanitization: There is no defined process for sanitizing, escaping, or validating the content retrieved from workspace files before it is interpolated into the agent's reasoning or written back to project records.

praxis-direction