Automate multi-layer self-review to prevent premature task completion in Rune.
npx skills add https://github.com/vinhnxv/rune-plugin --skill inner-flame
Please help me install this Agent Skill. Skill: inner-flame Source: https://github.com/vinhnxv/rune-plugin/tree/main/plugins/rune/skills/inner-flame Command: npx skills add https://github.com/vinhnxv/rune-plugin --skill inner-flame
The Inner Flame establishes a rigorous, role-adapted self-review that prevents premature task completion by enforcing evidence-backed validation and multi-layer checks.
Run the Inner Flame self-review protocol on the task you are about to seal.
High-intent search queries and answers about installing and using this skill.
To prevent premature task completion, apply a structured self-review protocol enforcing evidence-backed validation and multi-layer checks before sealing any task. This ensures outputs are grounded, complete, and verified against fresh evidence.
Grounding in self-review is a validation mechanism that prevents hallucination by ensuring sources are actually read and referenced. You need this process when traceable evidence and strict adherence to protocol are required for task reliability.
Perform self-adversarial review by running role-adapted checks across worker, fixer, reviewer, and researcher roles to surface weak findings. This structured process improves reliability by challenging conclusions during the execution and seal phases.
Yes, role-adapted checks can be applied across worker, fixer, reviewer, and researcher roles during planning, execution, and seal phases. This ensures completeness and correctness validations are tailored to the specific responsibilities of each role.
The best way to validate task completeness before sealing is implementing a three-layer review protocol: Grounding, Completeness, and Self-Adversarial. This multi-layer approach enforces fresh-evidence governance and produces seal-ready outputs.
Automated task output lacks traceable evidence when a rigorous self-review protocol is not enforced. Implementing grounding checks ensures sources are read and referenced, preventing hallucination and guaranteeing adherence to protocol.
dtsongTransform natural language project descriptions into structured engagement briefs.
jchildreeVerify completion claims by executing and confirming command outputs before commits.
KrishnatejavepaReview non-trivial code or design decisions with a fresh-context adversarial process.
gabriellpequenoRun lint, type checks, and validations for Node.js/TypeScript and Python projects.
