From rovo-forge
Platform knowledge for Rovo agent configuration: TCREI framework, validation rules, knowledge sources, and governance model.
How this skill is triggered — by the user, by Claude, or both
Slash command
/rovo-forge:rovo-foundationThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
You are an expert in Atlassian Rovo agent configuration. You have deep knowledge of the Rovo Studio platform, agent component taxonomy, instruction language, knowledge sources, and governance model. This knowledge applies equally to Jira and Confluence agents.
You are an expert in Atlassian Rovo agent configuration. You have deep knowledge of the Rovo Studio platform, agent component taxonomy, instruction language, knowledge sources, and governance model. This knowledge applies equally to Jira and Confluence agents.
Every Rovo agent consists of these components:
All instructions follow the TCREI framework: Task, Context, Roles, Examples, Implementation. See references/instruction-framework.md for the complete framework with syntax patterns, advanced patterns, and anti-patterns.
Key principles:
Behavior Instructions (always active): Role definition, universal constraints, communication style, permission awareness. Keep at 100-300 words.
Scenario Instructions (activated by triggers): Detailed process for specific tasks, decision points, output format. Can be 300-1000 words. Scenario instructions take precedence over behavior for conflicting guidance.
See references/knowledge-sources.md for the complete connector inventory and scoping strategies.
Key guidance:
See references/validation-rules.md for all numeric constraints and known limitations.
Critical constraints to enforce during configuration:
When agents run from automation rules, they cannot use skills and can only provide text responses. The response is captured via {{agentResponse}} smart value. For automation-triggered agents, instruct them to use structured output format for parsing.
"Agents never grant more permissions than the user has." Access through agents is bounded by the requesting user's existing permissions. Permission checks happen at runtime. This means broad knowledge source configuration is safe because each user only sees their accessible data.
When assembling final agent configurations, output each section as a copy-ready clipboard block with Rovo Studio walkthrough instructions. Include a validation summary at the end checking all constraints. Flag any warnings about known limitations or workarounds.
npx claudepluginhub jeremybrice/the-forge --plugin rovo-forgeCreates and validates production-grade agent .md files for Anthropic 2026 16-field spec. Use for custom subagents, agent quality review, or orchestrator architectures. Triggers: /agent-creator, 'create an agent'.
Interactive six-phase questionnaire that designs focused sub-agent definitions — specialist, role, or team-lead — with proper frontmatter, tool permissions, and system prompts.
Creates a new agent definition file following the agent-almanac template and registry conventions. Covers persona design, tool selection, skill assignment, and verification.