The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted data from GitHub issue titles, bodies, and comments to automate task decomposition, priority assignment, and agent coordination.
Ingestion points: The skill uses gh issue view, mcp__github__get_issue, and mcp__github__list_issues to pull external content into the agent's context (SKILL.md).
Boundary markers: There are no specific delimiters or instructions provided to the agent to distinguish between its own logic and potentially malicious instructions embedded in the issue content.
Capability inventory: The skill utilizes powerful capabilities including the Bash tool for shell execution, Write for file system modifications, and mcp__claude-flow__agent_spawn for creating new agent instances.
Sanitization: The prompts lack explicit sanitization or validation logic to prevent the agent from being influenced by instructions contained within processed GitHub issues.
[COMMAND_EXECUTION]: The skill heavily relies on the Bash tool to perform operations such as issue manipulation, project synchronization, and swarm management.
Evidence: The skill instructions involve executing complex shell scripts that pipe data between the gh CLI, jq, and other system utilities.
[EXTERNAL_DOWNLOADS]: The skill frequently executes remote code by invoking a package via npx.
Evidence: The skill uses npx ruv-swarm across multiple coordination tasks.
Note: The ruv-swarm package is associated with the skill's author (ruvnet) and represents intended vendor functionality.

agent-swarm-issue