# DenisSergeevitch/agents-best-practices

Provider-neutral Agent Skill for Codex, Claude Code, and agentic harness design.

Repository: https://github.com/DenisSergeevitch/agents-best-practices
Canonical: https://ross.abutalabs.com/products/agents-best-practices
License: MIT
License Family: permissive
Topics: agent-skills, agentic-workflows, ai-agents, codex, mcp, prompt-engineering, agents, codex-skill, agent-skill, anthropic, claude, claude-code
Last push: 2026-08-26T16:46:59+00:00

## Health v2 (maintenance only)
Score: 58/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 99, release rhythm 35, longevity 7
- inputs: {"age_days": 110, "days_push": 7, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, young
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2243, forks 204 (observed 2026-08-28T04:06:29.920125+00:00)

## What it is
A provider-neutral Agent Skill package that guides designing, auditing, refactoring, and explaining agentic harnesses for tools like Codex and Claude Code. It is installed into an agent's skills directory and provides best-practice knowledge rather than executable software.

## Use cases
- design an agentic harness for an AI agent
- audit and refactor my agent's tool permissions
- generate an MVP blueprint for a coding agent
- learn best practices for building AI agent workflows
- add agent skills to Claude Code or Codex
- understand planning mode and speculative tool execution in agents

## When to choose
- you are building or improving agentic workflows and harnesses
- you use Claude Code, Codex, or a compatible skills-based agent
- you want provider-neutral guidance on agent architecture, permissions, and observability

## When to avoid
- you need a runnable agent framework or runtime, not documentation
- your project does not support the skills directory convention
- you want a turnkey agent product rather than design guidance

## Facets
- artifact type: learning-resource
- maturity: active
- function: agent-framework, prompt-engineering, mcp, developer-tools
- domain: large-language-models, developer-tools
- platform: cli, cross-platform
- tags: agent-skills, claude-code, codex, agentic-harness, skill-package, best-practices, ai-agents, automation

## Member repositories
- DenisSergeevitch/agents-best-practices (main) score 58

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:06:29.920125+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T02:44:02.924335+00:00, confidence not recorded.
  - readme: https://github.com/DenisSergeevitch/agents-best-practices (fetched 2026-08-28T04:06:29.920125+00:00, sha 2badd91f3689)
- Data as of 2026-08-30T08:39:29.467469+00:00.