The Agent Skills Directory

[PROMPT_INJECTION]: The skill implements a pattern for a job queue that processes data from a database (jobs table). This creates a potential surface for indirect prompt injection if an attacker can insert or modify rows in the jobs table.
Ingestion points: Data enters the agent's context through the jobs table, specifically the type and args columns which determine which handler is run and with what arguments (observed in SKILL.md and references/sql-job-queue-extras.md).
Boundary markers: No specific delimiters or warnings against processing embedded instructions within job arguments are included in the provided templates.
Capability inventory: The skill possesses the capability to query the database, perform atomic updates, and dispatch task handlers (seen in the claim, tick, and dispatch functions).
Sanitization: The provided code snippets do not include explicit sanitization or validation of the args payload before processing, leaving this responsibility to the implementation of the handlers.
[SAFE]: No obfuscation, data exfiltration, or unauthorized command execution patterns were detected. The skill uses standard, well-known libraries and practices for database management and job scheduling.

sql-job-queue