The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data which could contain malicious instructions designed to influence the agent's behavior during the triage process.
Ingestion points: The skill takes an inbox item or batch to triage as its primary input (defined in SKILL.md).
Boundary markers: There are no explicit instructions or delimiters defined to separate the content of the inbox items from the agent's instructions, increasing the risk that the agent may follow instructions embedded within the data.
Capability inventory: The skill is authorized to update file metadata, move or rename files, and delete items from the inbox (defined in SKILL.md and references/canon-maintenance-contract.md).
Sanitization: There is no mention of sanitization, validation, or escaping of the content found within the inbox items before it is processed.

inbox-triage