AgentIndex icon
AgentIndex
ToolsMCP ServersHOTCategoriesTrendingNewBest ListsCompareGuides
Submit Tool
ToolsMCP ServersHOTCategoriesTrendingNewBest ListsCompareGuides
Home/
Code Assistant/
gentle-shell
gentle-shell logo

gentle-shell

Active·★ 2.8k·MIT·Updated 2026-11-17

gentle-shell turns a session of the Pi coding agent into a workspace with agent orchestration, live changes and runtime status, usage meters for supported provider accounts and built-in diff views. Its everyday workflow, Organic Driven Development, keeps questions, reviews and proposals read-only until you ask for a change, explores before editing, and gives larger authorized work one recoverable feature document so progress survives an interruption.

Focused explorer, writer and verifier sub-agents are only used for a stated reason, such as a high-risk change getting an independent verifier, and the README publishes small single-run benchmarks of how this routing cut time and cost. Other pieces include Engram memory, background jobs and monitors, Normal, Auto or YOLO permission modes, cross-session coordination and Gentle Design for HTML design directions. v4.1 added clean-machine installers for macOS, Windows and Linux. MIT-licensed; it requires Pi and is opinionated about workflow.

This write-up is based on the project's README and public repository data. We have not hand-tested the tool. Data as of 2026-11-13.

License
MIT
Latest release
v4.7.0 · 2026-11-07
Last commit
2026-11-13
Contributors
265
Forks
282
Open issues and PRs
516
First published
2026-05-10
npm package
gentle-pi 4.7.0 · 13,354 downloads last week

Source: GitHub, npm and PyPI public APIs.

#coding-agent#pi#terminal#code-review#memory#workflow
↗ Visit site★ GitHub
01

Features

01ODD workflow: read-only until a change is requested
02Focused explorer, writer and verifier sub-agents
03Native review of the exact diff
04Background jobs and output monitors
05Normal, Auto and YOLO permission modes
02

Why choose it

+Stays read-only until you ask for a change, so questions never turn into edits
+Sub-agents only when there is a reason, which the README shows cut time and cost
+Feature documents let long tasks resume after interruptions or compaction
03

Trade-offs

!Only for the Pi agent; not a plug-in for other coding agents.
!Published benchmarks are single runs (n=1).
!Opinionated workflow with many commands takes time to learn.
04

Maintenance check

ActiveChecked: 2026-11-12
Last commit
2026-11-12
Commits, last 90 days
100+
Releases, last 90 days
30
Latest release
v4.7.0 (2026-11-07)
Open issues
501
Recent issues with at least one comment
3/14
Commits per week, last 12 weeks
Commits per week, last 12 weeks: 40, 87, 197, 124, 226, 220, 254, 195, 64, 113, 277, 100

Oldest week on the left; the latest four weeks are orange. Source: GitHub commit activity.

Status rule: archived on GitHub = Archived; last commit within 90 days = Active; 91 to 365 days = Slowing; older = Stale. All figures come from the GitHub API on the date shown.

Most discussed open issues

  • #1064 feat(profiles): support session-bound active profiles without global routing mutation63 comments · 2026-09-15
  • #1154 bug(rdd): gentle_review_capture rejects fresh status collectBinding with capture-binding-rejected24 comments · 2026-09-17
  • #992 bug(review): committed-range START unreachable (3 contradictory prerequisite errors → native schema-incompatible) and inspect bases the rang23 comments · 2026-09-13
  • #741 bug(review): correction-plan captures always rejected — binding target (frozen authority) compared against live STATUS target21 comments · 2026-09-08
  • #2740 feat(design): canvas review loop — redesign, sizes and export, token inspector, history, mix and partial handoff, Spanish20 comments · 2026-11-02

Issue titles are quoted from GitHub. Open the link to check whether a problem still applies to the version you use.

05

Use cases

↳Pi users wanting a structured coding workflow with memory and review
↳Keeping long agent tasks resumable across sessions
06

Alternatives

⚡Top VS:
VSoh-my-pi★39.3k→VSCodewhale★41.2k→VSqwen-code★30.7k→VSponytail★193.0k→
oh-my-pi logo
oh-my-pi★ 39.3k
⌥ Coding agent with the IDE wired in. Built by Stencil Labs.
VSCompare→
Codewhale logo
Codewhale★ 41.2k
Open-source agent for the terminal written in Rust. Use any provider or model. Issues & PRs always welcome
VSCompare→
qwen-code logo
qwen-code★ 30.7k
An open-source AI coding agent that lives in your terminal.
VSCompare→
ponytail logo
ponytail★ 193.0k
Makes your AI agent think like the laziest senior dev in the room. The best code is the code you never wrote.
VSCompare→
jcode logo
jcode★ 21.9k
High performance coding agent harness written in Rust, built for heavy multi-agent dev workflows
VSCompare→
dsh-Go logo
dsh-Go★ 26.0k
dsh-go — DeepSeek Harness 装进口袋:手机本机运行,也能遥控你的电脑。
VSCompare→
PI-Desktop logo
PI-Desktop★ 8.0k
Local-first AI coding agent desktop: Electron + Rust host core + pi Agent Harness + user-installable plugins
VSCompare→
Kun logo
Kun★ 6.4k
Local-first AI agent workspace for coding, writing, design, research, and automation — one runtime for desktop GUI and TUI.
VSCompare→
See all alternatives →

Comments

Log in to leave a comment

No comments yet. Be the first!

On this page
01Features02Why choose it03Trade-offs04Maintenance check05Use cases06Alternatives
Gentleman-Programming /
gentle-shell
↗

Coding-agent workspace on Pi with a read-first workflow, focused sub-agents, memory and native diff review

2.8k282MIT
Last commit1d ago
StatusActive
CategoryCode Assistant
Contributors265
Latest releasev4.7.0
Links
Documentation↗Discussion↗Issues↗Releases↗
Cloud Infrastructure
Ad

Deploy on DigitalOcean

Host your MCP servers, Docker containers, and AI agents with $200 free credit over 60 days.

Claim $200 Free Credit
© 2026 AgentIndex.app|Built by a 10-year iOS Developer.
Best by use caseAboutContactAdvertisePrivacy PolicyTerms of UseQYSGitHubBuy me a coffee ☕

Browse by Category

MCP ServersCode AssistantWorkflow AutomationRAG / Knowledge BaseMulti-AgentBrowser AutomationLLM InfraDev ToolingObservability

Not affiliated with Anthropic, OpenAI or Microsoft.