The Agent Skills Directory

[PRIVILEGE_ESCALATION]: The skill instructions direct the agent to re-run commands with escalated permissions if a runtime_access_denied error occurs, which encourages the acquisition of higher privileges to bypass sandbox restrictions.
[INDIRECT_PROMPT_INJECTION]: The skill processes data from external Linear tickets, comments, and attachments, creating a potential vector for indirect prompt injection attacks.
Ingestion points: Linear ticket text, comments, and attachments (SKILL.md).
Boundary markers: The skill includes an explicit instruction to treat this data as untrusted and never as instructions, though no technical delimiters (like XML tags or JSON schemas) are enforced in the prompt.
Capability inventory: Shell command execution via the orca CLI tool (SKILL.md).
Sanitization: No specific technical sanitization or escaping mechanisms are described for the external content before it is processed.
[COMMAND_EXECUTION]: The skill relies on shell command execution via local binaries (orca, orca-ide, orca-dev) to interact with the Linear platform.

linear-tickets