The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes GitHub CLI (gh) commands to search for PRs and issues and to create new ones. These actions are restricted to the pollinations/pollinations repository as per the hard rules.- [PROMPT_INJECTION]: The skill has a surface for indirect prompt injection as it reads existing GitHub content to inform its output. 1. Ingestion points: SKILL.md (Workflow Step 1) via gh search prs and gh search issues. 2. Boundary markers: Absent; there are no specific instructions to ignore malicious directives within the retrieved data. 3. Capability inventory: gh issue create and gh issue comment are available tools. 4. Sanitization: Absent; the fetched content is used to influence the agent's style without explicit filtering. The risk is low given the narrow focus of the skill.

issue-maker