Analyze MoonMind task runs and append a validated follow-up task proposal.
npx skills add https://github.com/MoonLadderStudios/MoonMind --skill fix-proposal
Please help me install this Agent Skill. Skill: fix-proposal Source: https://github.com/MoonLadderStudios/MoonMind/tree/main/.agents/skills/fix-proposal Command: npx skills add https://github.com/MoonLadderStudios/MoonMind --skill fix-proposal
Detects execution quality problems in MoonMind task runs and generates a single, high-value follow-up task proposal to address the root cause, reducing repeated failures and wasted retries.
Use the fix-proposal skill to analyze the latest task run and append one actionable proposal to the task_proposals.json file.
High-intent search queries and answers about installing and using this skill.
To generate fix proposals for task run failures, analyze task_context.json and run artifacts to detect degraded behaviors like retry storms and missing references, appending a validated taskCreateRequest to the proposals JSON.
Execution quality issues detected from MoonMind run logs include access failures, conflicting instructions, execution loops, duplicate output, missing references, and retry storms that degrade task run behavior.
Append a follow-up task proposal to task_proposals.json by generating one validated proposal object with a unique title, summary, severity signal, and taskCreateRequest while preserving valid JSON formatting and avoiding duplicate titles.
Yes, you can detect retry storms and access failures by analyzing task run artifacts and logs, generating a single high-value proposal object with category tags and evidence to address the root cause.
The best way to prevent repeated task failures and wasted retries is to analyze run logs for degraded behaviors and produce a targeted fix proposal that adjusts retry logic and resolves missing references.
Yes, you need task_context.json alongside artifacts and logs to accurately diagnose execution loops and duplicate output, ensuring the generated fix proposal contains valid evidence and a corrective taskCreateRequest.
MoonLadderStudiosAnalyze task context and logs to propose one continuation task with metadata.
hungtrandigitalGuide root-cause investigation through a four-phase systematic debugging workflow.
mec07Delay execution for specified durations and trigger follow-up skills.
tunnox-netAudit authentication, encryption, and governance for security vulnerabilities.
