DenisSergeevitch/agents-best-practices
This skill provides a technical blueprint for building agentic systems that remain compatible across different LLM providers. It focuses on the structural requirements of agent loops, from tool design and record provenance to memory lifecycles and context compaction.
It is designed for builders auditing or refactoring agent architectures to improve observability, safety, and the implementation of self-refining recursive harnesses.
Claude
CodexUse this skill
Follow the documented setup, then try a first task.
Global install with skills
Installs: Global, user-level installation discoverable by every project · any compatible agent
Run this command in a terminal. The documented -g flag installs the skill globally at user level.
npx skills add DenisSergeevitch/agents-best-practices -g
Ask Codex to install it
Installs: User-level agents-best-practices skill directory · Codex
Enter the following request in a Codex agent session, not in a terminal: “Install the agents-best-practices skill for me: 1. Clone https://github.com/DenisSergeevitch/agents-best-practices into my user-level skills directory as `agents-best-practices/`. Use the skill directory my agent reads on this machine, for example: Codex: ~/.codex/skills/ 2. Verify that SKILL.md, icon.jpeg, and the references/ directory are present. 3. Confirm the install path when done.”
Ask Claude Code to install it
Installs: User-level agents-best-practices skill directory · Claude Code
Enter the following request in a Claude Code agent session, not in a terminal: “Install the agents-best-practices skill for me: 1. Clone https://github.com/DenisSergeevitch/agents-best-practices into my user-level skills directory as `agents-best-practices/`. Use the skill directory my agent reads on this machine, for example: Claude Code: ~/.claude/skills/ 2. Verify that SKILL.md, icon.jpeg, and the references/ directory are present. 3. Confirm the install path when done.”