The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection because it processes untrusted data from the PagerDuty API that could influence agent behavior, while also providing tools to modify data.
Ingestion points: Tools such as list_oncalls, get_schedule, list_schedules, list_schedule_users, and get_escalation_policy in SKILL.md fetch external content into the agent context.
Boundary markers: No delimiters or instructions are present to prevent the agent from following instructions potentially embedded in schedule names, notes, or policy details.
Capability inventory: The skill provides significant write access via create_schedule, update_schedule, and create_schedule_override tools in SKILL.md.
Sanitization: There is no evidence of sanitization or validation logic to filter potentially malicious strings from API responses.
[NO_CODE]: The skill is composed strictly of configuration and markdown instructions in SKILL.md and does not include any accompanying source code or scripts for execution.

pagerduty-oncall