@epanonymouschecks…add-bitplanet-collaboration-skills → master1 files · +2 −0updated 4mo agoExtracts self-improvement insights from agent session history, creating dated changelogs of mistakes and corrections.
What problem it solves: Agents make repeated mistakes because learnings aren't systematically captured. This skill creates a structured changelog of agent corrections that future agent instances can learn from.
Who uses this: Anyone running persistent AI agents who wants to improve agent behavior over time.
GitHub: https://github.com/Bitplanet-L1/agent-skills/tree/main/agent-learnings
Analyzes interaction patterns to produce guides for humans on working effectively with AI agents.
What problem it solves: Humans struggle to communicate effectively with agents, leading to repeated frustrations. This skill extracts patterns from real interactions to document what works and what doesn't.
Who uses this: Anyone learning to work with AI agents, or teams onboarding new members.
GitHub: https://github.com/Bitplanet-L1/agent-skills/tree/main/human-agent-collaboration
Both skills emerged from real operational experience running AI agents. They capture patterns about agent cognition, verification discipline, memory limitations, and effective communication.
Attribution: Built by Bitplanet, powering Deva (https://deva.me)