#8282nd of 98 in Skills, plugins and rules
gentle-ai
Go installer adding memory, workflow and review rules to 17 coding agents
Documentation ↗
(opens in a new tab)Website ↗
(opens in a new tab)Repository on GitHub ↗
(opens in a new tab)
- Stars
- 7.6k
- License
- MIT
- Last commit
- Oct 2026
- Last release
- Oct 2026
- Language
- Go
Missed the latest GitHub check; these facts may be older.
Overview
Gentle-AI is a Go CLI that configures coding agents you already have (Claude Code, Codex, OpenCode, Cursor, Pi and 12 more) with persistent memory (Engram), skills, optional Context7 MCP and a credential-file deny-list. It adds two workflows: ODD, which sizes work and decides when to delegate to subagents, and RDD, where the binary freezes a change and sets review depth by risk. Configs are snapshotted before each write.
Who it is for: Developers on Claude Code, Codex, OpenCode or Cursor wanting shared memory
Strengths
- Configures 17 agents using each agent's native configuration
- Config backups before every write, plus a read-only `gentle-ai doctor` check
- Review transitions come from the binary and on-disk state, not model judgment
- Security deny-list blocks ~/.ssh, .env and credential files
Weaknesses
- Does not install an agent; requires one already set up
- Delegation and RDD review features differ per agent
- Benchmarks are single runs (n=1) on the authors' own suite
- Windows install is a Go source build needing Go 1.25.10+
What it needs
- no GPU
- Needs An existing supported AI coding agent, Go 1.25.10+ (Windows source install), Engram, Context7 MCP (optional)
Also in Skills, plugins and rules
See all 98| Rank | Project | Score |
|---|---|---|
| 1 | skillsAgent skills for planning, TDD, debugging and code review | 94 out of 100 |
| 2 | eccAgents, skills, hooks and rules bundle for Claude Code, Codex and other harnesses | 94 out of 100 |
| 3 | cavemanCuts token use in coding agents with terse replies and compressed inputs | 94 out of 100 |
| #4 | impeccableDesign skill with 24 commands and a detector for AI-generated frontends | 93 out of 100 |
| #5 | ponytailSkill prompt that makes coding agents write less code and add tests | 92 out of 100 |