 [](https://github.com/PiLastDigit/TRIP-workflow/blob/master/LICENSE)  [](https://docs.anthropic.com/en/docs/claude-code) [](https://developers.openai.com/codex/cli/) [](https://github.com/sst/opencode) [](https://github.com/mistralai/mistral-vibe)
## What is TRIP?
A structured development workflow for AI coding agents that brings **memory**, **consistency**, and **reduced hallucination** (only humans should) to AI-assisted development. TRIP helps you enter flow state and eat features like buttered noodles.
It is also the acronym (reversed) of the historical 4-phases development cycle: **P**lan, **I**mplement, **R**eview, **T**est.
**Note:** Since v2.0.0 the flow is even simpler **Plan → Implement → Release** — review and test moved *inside* Implement as a testing gate and an automatic Codex review loop, every feature passes through all 4 phases with fewer commands.
TRIP was initially designed for Claude Code using the [Agent Skills](https://agentskills.io/home) open standard (`SKILL.md`). Also compatible with OpenCode, Codex CLI, Mistral Vibe and more.
## Why TRIP?
There are tons of AI coding workflows out there like [Superpowers](https://github.com/obra/superpowers), [BMAD](https://github.com/bmad-code-org/BMAD-METHOD), [Gastown](https://github.com/steveyegge/gastown) and countless others. They might be powerful, but overwhelming for many of us dumb asses.
Even the "simple" ones come with:
- 47 different commands & skills to memorize
- Sub-agents swarm for God-knows-what
- Multi-chapters courses (sometimes paid lol)
**TRIP is different.** It's deliberately minimal:
| That's it | Just these |
| ------------------- | ------------------------------------------------------ |
| `/TRIP-1-plan` | Think before you code |
| `/TRIP-2-implement` | Implement, test & review |
| `/TRIP-3-release` | Version, changelogs, docs, commit, tag, merge, push |
3 numbered skills. 1 architecture file. 0 PhD required.
The onboarding is: copy the folder, run init, start coding. If you can count to 3, you can TRIP.
It was kept stupid simple because **the goal is to ship features, not to master a workflow**. The workflow should disappear into the background, not become a project of its own.
## Getting Started
1. Copy the `skills/` folder contents to your repo's `.claude/skills/` or whatever
2. Run `/TRIP-init [YourProjectName]`
3. Follow the interactive prompts
4. Review and approve the generated ARCHI.md
### Additional For Mistral users (if they exist)
Also copy `AskUserQuestion/` to your agent `/skills/`, it provides the `AskUserQuestion` hook that TRIP workflow rely on.
Et voila ! Start using the skills like `/TRIP-1-plan auth for this webapp`, `/TRIP-2-implement @auth-plan.md`, etc.
By default `/TRIP-1-plan` chains straight into implementation once the Codex plan review approves: answer the discovery questions, walk away, and come back to an implemented feature. Want to read the plan before any code is written? Add `--review`. `/TRIP-2-implement` still runs on its own for a plan written earlier or a resumed session.
https://github.com/user-attachments/assets/d37bbc60-1868-4fa8-9be6-083b60d6a53d
## The Heart of TRIP: ARCHI.md
The `ARCHI.md` file is the **central nervous system** of this workflow. It serves as the AI agent's **long-term memory** of your codebase.
### Why ARCHI.md Matters
**1. Persistent Context Across Sessions**
AI agents have no memory between sessions. Every new conversation starts from zero. ARCHI.md solves this by providing a comprehensive, always-up-to-date snapshot of your architecture that the agent reads at the start of each task. Unlike tool-specific files like `CLAUDE.md` or `AGENTS.md`, ARCHI.md is purely about architecture. It's tool-agnostic, so it works with any agent. You can still reference it from your `CLAUDE.md` to include it in all conversations.
**2. Token Savings & Reduced Hallucination**
Without ARCHI.md, your agent must glob, grep, and read multiple files to piece together the architecture from scratch for every single session. This wastes tokens and leads to guessing: _"There's probably a utils folder..."_, _"This project likely uses Redux..."_. ARCHI.md eliminates both problems. The agent gets the full picture in one read for minimal exploration & hallucination.
**3. Balanced Detail vs Token Usage**
ARCHI.md is designed to be:
- **Detailed enough** to provide meaningful context, **concise enough** to not waste tokens
- **Structured** for quick navigation
- **Updated** after every architectural change
It's not a dump of your entire codebase, rather a curated architectural guide.
## The Init Process
The `TRIP-init` skill is a **script written in human language** that programmatically bootstraps the TRIP workflow in any repository.
### What Init Does
1. **Creates the docs structure** - Folders for plans, changelogs, reviews, tests, memos
2. **Explores your codebase** - Identifies languages, frameworks, patterns, conventions
3. **Classifies your project** - Web frontend? CLI tool? Embedded firmware? Library?
4. **Generates ARCHI.md** - Tailored to your specific project type
5. **Customizes the skills** - Replaces placeholders with your project's specifics
### The Placeholder System
The generic TRIP skills contain placeholders like:
- `[PROJECT_NAME]` - Your project's name
- `[VERSION_FILE]` - Where your version is stored (package.json, Cargo.toml, etc.)
- `[ADAPT_TO_PROJECT: ...]` - Sections to customize
Init walks you through questions and replaces these placeholders based on your answers, creating a workflow tailored to your project.
## More Skills
### `/codex-implement`
Implementation delegated to Codex CLI in a workspace-write sandbox.
### `/codex-plan-review` & `/codex-code-review`
Iterative review loops powered by Codex CLI. The code review checklist and the code-review record template live in `codex-code-review/` and are adapted to your project by init.
### `/codex-ask`
A grounded second opinion on **anything**. TRIP-research uses it to brainstorm findings before presenting them.
### `/TRIP-upgrade`
Upgrades an existing project's TRIP skills to the latest release without losing project customizations. Run `/TRIP-upgrade`, done: it fetches this repository itself (pass a version to pin one, or a local folder to upgrade from a fork). Under the hood it recovers the generic template you installed from, computes your customizations as a diff, and replays them onto the new version with a git three-way merge, so the result is exact and every change is shown as a diff before it lands.
**Coming from a version before 2.8.4** (no self-fetching upgrade skill yet)? One-time bootstrap: copy this repo's `skills/TRIP-upgrade/` over your `.claude/skills/TRIP-upgrade/`, then run `/TRIP-upgrade`. Installs older than 2.8.1 carry no version stamp, so that first run brings them to 2.8.4 by hand, then merges them to the latest release; every run after that merges directly.
Not sure which TRIP version a project runs? Every skill carries it in its frontmatter (`metadata.trip-version`, since v2.8.1):
```bash
grep -h "^ trip-version" .claude/skills/*/SKILL.md | sort | uniq -c
```
One line back means a clean install; several lines mean some skills were skipped in a past upgrade. No output means a pre-2.8.1 install.
### `/TRIP-hotfix`
Streamlined workflow for production emergencies. Bypasses full TRIP for genuine crises (or lazy debugging).
### `/TRIP-research`
Exploratory investigation with defined compute level. For feasibility studies and technology evaluation. Produces documented findings, not production code.
### `/TRIP-compact`
Run this skill to compact ARCHI.md size while preserving relevance, accuracy, and coverage through summarization and restructuring. Token calculator script included. `/TRIP-compact --sync` skips compression and audits ARCHI.md against the actual codebase instead, reporting stale claims and gaps before fixing them.
## Multi-Agent: Using Different LLMs at Different Steps
Just like you wouldn't smell your own fart, an LLM is unlikely to catch bugs in its own implementation. Some people conduct adversarial review with a different session but still the same model, which is..._meh_. The best approach is to introduce a different model in the same reasoning ballpark as the first one, that will most likely catch what the other missed.
As of v2.0.0, this multi-agent approach is **the default workflow**.
Considering Claude Code as your main agent and Codex as the copilot, the recommended approach to save tokens is to split in two sessions.
Session 1 with Fable for planning, Session 2 with Opus for implementation (or switch model mid-session).
A default `/TRIP-1-plan` run implements on the same model that planned. To split, run `/TRIP-1-plan <feature> --review`, pick "Approved — stop here", then `/TRIP-2-implement <plan>` in the Opus session.
```mermaid
flowchart TD
A["<b>/TRIP-1-plan</b> · Fable<br/>Discovery and plan draft"] --> B{"ChatGPT Astra<br/>plan review"}
B -->|"REQUEST_CHANGES"| Bf["Fable fixes the plan"]
Bf -->|"re-review"| B
B -->|"APPROVED"| D["<b>/TRIP-2-implement</b> · Opus<br/>Branch + split<br/>to-dos into batches"]
Bf ~~~ D
D --> E["ChatGPT Luna<br/>implements a batch"]
E --> F["Opus reviews the delta,<br/>fixes directly"]
F -->|"next batch"| E
F -->|"all batches done"| G["Opus final pass<br/>+ testing gate"]
G --> H{"ChatGPT Sol<br/>full code review"}
H -->|"REQUEST_CHANGES"| Hf["Opus fixes + re-tests"]
Hf -->|"re-review"| H
H -->|"APPROVED"| K["<b>/TRIP-3-release</b><br/>Version bump · changelog<br/>docs/ARCHI update · docs sync<br/>commit · tag · ff-merge · push"]
Hf ~~~ K
```
As of late September 2026, this Fable + Opus + GPT-6 harness combo is absolute peak.
Recommended plans: Claude Max 5x ($100) + ChatGPT Plus ($20). The compute value you get out of those 2 is still unmatched, no chinese API can compete with those subsidized subs. Enjoy while it lasts.
## MCP Servers: Less Is More
Last piece of advise before your new coding quest: Every MCP server you add is extra context, extra latency, and extra confusion. Keep it minimal. The one use case where MCP genuinely shines is **up-to-date documentation**, so your agent stops hallucinating deprecated APIs/whatever. Two servers cover it: [Context7](https://github.com/upstash/context7) for current library & framework docs, and [Exa](https://github.com/exa-labs/exa-mcp-server) for web search when the answer isn't in any doc. No bloat beyond that.
## Contributing
PRs & forks are welcome
Happy tripping ! 🍄