The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill operates on a shared knowledge graph ("methodology" realm), reading existing principles and creating new ones, which allows for potential instruction injection from the graph data.
Ingestion points: The skill ingests data from the methodology graph using the iskron_orient and iskron_search tools as part of its orientation and search workflows (SKILL.md).
Boundary markers: The instructions lack specific boundary markers or "ignore embedded instructions" warnings for the agent when processing content retrieved from the graph.
Capability inventory: The skill possesses search, read, and write capabilities via iskron_search, iskron_look, and iskron_case (SKILL.md).
Sanitization: There are no explicit instructions to sanitize or validate graph content before it is interpolated into the agent's context or used to generate new methodology nodes.

methodology-work