The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and act upon data retrieved from an external graph system, creating a surface for indirect instructions to influence agent behavior.
Ingestion points: The agent retrieves node descriptions, attributes, and relationship data using the iskron_look and iskron_orient tools (e.g., in SKILL.md sections describing the weave and reverify groups).
Boundary markers: The skill does not explicitly define delimiters or specific instructions to isolate and ignore potentially malicious directives embedded within the node content being processed.
Capability inventory: The agent possesses write-access capabilities to the graph via the iskron_arrow tool (actions link, update, reconnect), allowing it to modify the data structure based on interpreted content.
Sanitization: There is no mention of sanitization, filtering, or validation of the semantic content (prose) retrieved from the graph before the agent uses it to make structural decisions.

weaving