Process human feedback to implement fixes and record learnings.
npx skills add https://github.com/victoriacity/openakari --skill feedback-victoriacity
Please help me install this Agent Skill. Skill: feedback Source: https://github.com/victoriacity/openakari/tree/main/.claude/skills/feedback Command: npx skills add https://github.com/victoriacity/openakari --skill feedback-victoriacity
This Skill allows for direct human feedback, corrections, and strategic direction to be applied to an agent's work, ensuring continuous improvement and alignment with human intent.
/feedback to correct the behavior, have the agent investigate why it failed, implement a code or prompt fix, and record the learning so the agent adheres to the protocol in the future.Use the feedback skill to correct the agent's recent code commit by stating 'Do not merge code without a passing test suite'.
High-intent search queries and answers about installing and using this skill.
You can correct agent behavior by providing direct instructions through human-in-the-loop feedback. The skill processes your corrections, investigates the root cause in code or history, implements fixes, and records the learning to prevent recurrence.
Yes, workflow adjustment is a core feature. You can inject directives to update project conventions, modify prompts, or shift the project roadmap, ensuring the agent aligns its operational behavior with your strategic intent.
The skill processes diverse feedback types including corrections, complaints, directives, observations, approvals, resource adjustments, strategy shifts, knowledge injection, quality calibration, tuning, and scheduling.
Yes, root cause analysis is performed by investigating code, history, and prior feedback. It identifies why the agent failed, implements actionable fixes via code or prompt adjustments, and records the cycle to prevent future recurrence.
The best way is using direct feedback to inject knowledge and calibrate quality. The skill interprets your directives, adjusts operational alignment through code or convention updates, and logs the resolution.
Provide a direct correction stating the required protocol. The agent investigates why it failed the deployment protocol, implements a code or prompt fix to align, and records the learning to ensure future adherence.
danzam98Review AGENTS.md and README.md to map project architecture and purpose.
ThalesGroupExecute targeted AGILAB tests across unit, smoke, and regression suites.
JoNaYeonAutomate code quality gates and security standards across project workflows.
mangowhoiscloudDebug autonomous agent systems using root cause analysis and ContextVar lifecycle patterns.