IgorGanapolsky
Are you IgorGanapolsky? Claim your skills.
IgorGanapolsky / thumbgate
ThumbGate enhances AI coding agents by capturing feedback and preventing repeated mistakes through structured rules and gates.
IgorGanapolsky / thumbgate
ThumbGate enhances AI coding agents by capturing feedback and preventing repeated mistakes through structured rules and gates.
IgorGanapolsky / rlhf-feedback
Captures and dual-writes feedback for analytics and memory recall across platforms, enhancing model training and preventing errors.
IgorGanapolsky / second-pass
Facilitates a second review of tasks by handing off to Codex for independent analysis within Claude Code.
IgorGanapolsky / capture-feedback
Captures structured feedback with context and tags after tasks, enhancing lesson distillation and improving future outcomes.
IgorGanapolsky / rlhf-feedback
Captures user feedback to enhance agent memory and improve responses based on explicit quality signals.
IgorGanapolsky / adversarial-review
Conducts thorough Codex reviews to identify risks and regressions before critical code changes, enhancing deployment safety.
IgorGanapolsky / review
Facilitates Codex reviews on uncommitted work or specific commits, enhancing code quality through detailed analysis.
IgorGanapolsky / setup
Validates the Codex bridge plugin setup in Claude Code, ensuring proper installation and configuration for Codex review.
IgorGanapolsky / rlhf-feedback
Captures user feedback to enhance agent performance through structured memories and prevention rules based on explicit quality signals.
IgorGanapolsky / prevention-rules
Automates the generation and review of prevention rules based on repeated failure patterns to enhance project reliability.
IgorGanapolsky / recall-context
Helps developers retrieve historical context and prevention rules to avoid past coding failures before starting new tasks.
IgorGanapolsky / search-lessons
Enables users to search a database for lessons on corrective actions, lifecycle states, and linked rules for effective problem resolution.
IgorGanapolsky / result
Retrieves the latest saved Codex bridge result from Claude Code, allowing users to inspect previous outputs without rerunning Codex.
IgorGanapolsky / status
Displays the latest Codex bridge artifact metadata, including mode, timestamp, and exit status for efficient debugging.
IgorGanapolsky / Agent Memory
Enables agents to recall past mistakes and capture feedback, enhancing performance by preventing repeated errors through local memory management.
IgorGanapolsky / solve-architecture-autonomy
Automates the handling of architectural issues, crisis management, and deployment errors to enhance operational efficiency.