The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from external issue trackers (GitHub, Linear, and Jira) to categorize tasks and route them to other agent skills. Maliciously crafted issue titles, bodies, or comments could attempt to influence the agent's behavior during the routing or the subsequent execution phase.
Ingestion points: External content is fetched using scripts/tickets.sh and scripts/map.sh and processed in the routing logic defined in references/board.md and references/issue.md.
Boundary markers: The skill does not employ explicit boundary markers or instructions to ignore embedded commands when presenting external issue content to the agent.
Capability inventory: The skill uses git and gh to interact with the repository and trackers, and it routes tasks to other skills which may possess broader system or network permissions.
Sanitization: Content from external trackers is ingested as raw text and is not sanitized before being analyzed by the agent.
[COMMAND_EXECUTION]: The skill relies on several local bash and python scripts to perform its operations, including interacting with remote APIs and managing local state.
Evidence: The workflow execution spine in SKILL.md invokes scripts/tickets.sh, scripts/map.sh, and scripts/run-state.sh to perform core tasks like listing issues and managing run locks. scripts/tickets.sh includes an embedded Python adapter for Linear and Jira integrations.

portal