Submit structured Phase 4 feedback with failure categories and evidence.
npx skills add https://github.com/myrtlepn/gran-maestro --skill feedback-myrtlepn
Please help me install this Agent Skill. Skill: feedback Source: https://github.com/myrtlepn/gran-maestro/tree/main/skills/feedback Command: npx skills add https://github.com/myrtlepn/gran-maestro --skill feedback-myrtlepn
This Skill allows users to provide structured, manual feedback on AI-generated output within the Gran Maestro workflow, specifically addressing issues in Phase 4 (feedback loop) to ensure continuous improvement and accurate implementation.
ac_unclear, interpretation, implementation) and provide specific evidence (logs, screenshots, metrics).interpretation failure, provide the relevant AC ID and a screenshot of the incorrect output, and request a re-evaluation based on the correct interpretation.Submit feedback for request PLN-001, indicating an implementation error with evidence from log file /path/to/log.txt.
High-intent search queries and answers about installing and using this skill.
To submit manual feedback on AI-generated output, categorize the failure as `ac_unclear`, `interpretation`, or `implementation`, then provide specific evidence like logs or screenshots linked to the relevant AC ID and specify rework instructions.
AI quality assurance refinement categorizes failures into three types: `ac_unclear` for ambiguous acceptance criteria, `interpretation` for misunderstood requirements, and `implementation` for execution errors, enabling automated routing to the correct rework protocol.
Yes, implementation errors support automated routing to external services like `/mst:codex` or `/mst:gemini` for re-executing tasks, ensuring that execution failures are automatically addressed through the appropriate external rework protocols.
The best way to provide targeted rework instructions is to focus on restoring specific acceptance criteria by linking evidence to AC IDs, guiding the AI to re-evaluate based on correct interpretations rather than explicitly dictating the underlying implementation details.
Yes, the feedback mechanism manages feedback round limits to ensure continuous improvement. It restricts the number of rework iterations within the workflow, preventing endless loops during the AI output refinement process.
mrvladd-dVerify task completion against acceptance criteria and generate verification reports.
myrtlepnResume interrupted AI workflows from saved file-based task states.
myrtlepnView and modify Gran Maestro settings in `.gran-maestro/config.json`.
droxeyGuide end-to-end coding workflows with tests-first implementation and mutation testing.