The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content such as issue comments, pull request titles, and code annotations which can be controlled by external third parties.
Ingestion points: Data is ingested from external issue trackers (GitHub, GitLab, Linear) and repository metadata files.
Boundary markers: The skill includes a dedicated '' section that instructs the agent to treat third-party text as data and not as instructions.
Capability inventory: The skill uses the Agent tool to spawn sub-workers and the Bash tool for repository-level operations.
Sanitization: It explicitly defines 'attack signals' for identifying instructional content within data and requires the agent to stop for human review if such patterns are detected.
[COMMAND_EXECUTION]: The skill executes a local shell script, templates/resolve-config.sh, to manage its configuration. This script uses standard utilities like git and jq to read and merge local settings from .tituskirch-skills.json.
[DYNAMIC_EXECUTION]: The skill uses the Agent tool to dynamically spawn independent sub-agent workers to process each issue in the queue, representing a core architectural delegation pattern.

work-review-queue