haizhongzheng · Nov 6, 2026 · Nov 6, 2026
diff --git a/README.md b/README.md
@@ -15,7 +15,7 @@
 [![WeChat](https://img.shields.io/badge/WeChat-community-07C160?logo=wechat&logoColor=white)](assets/community/wechat.md)
 [![Discord](https://img.shields.io/badge/Discord-community-5865F2?logo=discord&logoColor=white)](https://discord.gg/CKjBjgVQDV)
 
-**[Get Started](#get-started)** · **[Harness Adapters](#what-is-a-harness-adapter)** · **[Methods](#supported-harness-evolution-methods)** · **[Examples and Recipes](#examples-and-recipes)** · **[FAQ](#faq)** · **[Docs](https://infini-ai-lab.github.io/hermit/docs/)**
+**[Get Started](#get-started)** · **[Harness Adapters](#what-is-a-harness-adapter)** · **[Methods](#supported-harness-evolution-methods-more-to-come)** · **[Examples and Recipes](#examples-and-recipes)** · **[FAQ](#faq)** · **[Docs](https://infini-ai-lab.github.io/hermit/docs/)**
 
 </div>
 
@@ -25,6 +25,7 @@ Hermit is a framework for training, evaluating, and deploying harness adapters,
 
 ## News
 
+- **[2026/11]** **Harness evolving with your coding plan. No API key needed.** Check out the [tutorial](docs/tutorials/codex-plan.md).
 - **[2026/11]** **Hermit v0.2.0** released — three-layer agents, Codex plan-backed acting and learning, and six new harness evolution methods. See the [release notes](https://github.com/Infini-AI-Lab/hermit/releases/tag/v0.2.0).
 - **[2026/10]** New harness evolution supported: **[Meta-Harness Adapter](https://infini-ai-lab.github.io/hermit/docs/methods/meta_harness_adapter.html)**, **[ReHarness](https://infini-ai-lab.github.io/hermit/docs/methods/reharness.html)**, and **[Harness²](https://infini-ai-lab.github.io/hermit/docs/methods/harness2.html)**.
 - **[2026/10]** **Hermit v0.1.0** released — the first public release for training, evaluating, and deploying harness adapters. See the [release post](https://x.com/haizhong_zheng/status/2114768121705775463).
@@ -43,22 +44,19 @@ Many harnesses are black boxes; even open-source ones have carefully designed co
 
 ## A Unified Framework for Harness Evolution
 
-### Supported Harness Evolution Methods
+### Supported Harness Evolution Methods (More to Come)
 
-<table align="center">
-<thead>
-<tr>
-<th>Status</th>
-<th>Methods</th>
-</tr>
-</thead>
-<tbody>
-<tr>
-<td>✅ <strong>Supported</strong></td>
-<td><a href="https://infini-ai-lab.github.io/hermit/docs/methods/gepa.html">GEPA</a> · <a href="https://infini-ai-lab.github.io/hermit/docs/methods/skillopt.html">SkillOpt</a> · <a href="https://infini-ai-lab.github.io/hermit/docs/methods/mwm.html">MetaWorld Model</a> · <a href="https://infini-ai-lab.github.io/hermit/docs/methods/harness2.html">Harness²</a> · <a href="https://infini-ai-lab.github.io/hermit/docs/methods/continual_harness.html">Continual Harness</a> · <a href="https://infini-ai-lab.github.io/hermit/docs/methods/reharness.html">ReHarness</a> · <a href="https://infini-ai-lab.github.io/hermit/docs/methods/meta_harness_adapter.html">Meta-Harness Adapter</a> · <a href="https://infini-ai-lab.github.io/hermit/docs/methods/ace.html">ACE</a> · <a href="https://infini-ai-lab.github.io/hermit/docs/methods/ahe_adapter.html">AHE Adapter</a> · <a href="https://infini-ai-lab.github.io/hermit/docs/methods/rrsi_adapter.html">RRSI Adapter</a> · <a href="https://infini-ai-lab.github.io/hermit/docs/methods/evosafe.html">EvoSafeHarness</a></td>
-</tr>
-</tbody>
-</table>
+- [GEPA](https://infini-ai-lab.github.io/hermit/docs/methods/gepa.html)
+- [SkillOpt](https://infini-ai-lab.github.io/hermit/docs/methods/skillopt.html)
+- [MetaWorld Model](https://infini-ai-lab.github.io/hermit/docs/methods/mwm.html)
+- [Harness²](https://infini-ai-lab.github.io/hermit/docs/methods/harness2.html)
+- [Continual Harness](https://infini-ai-lab.github.io/hermit/docs/methods/continual_harness.html)
+- [ReHarness](https://infini-ai-lab.github.io/hermit/docs/methods/reharness.html)
+- [Meta-Harness Adapter](https://infini-ai-lab.github.io/hermit/docs/methods/meta_harness_adapter.html)
+- [ACE](https://infini-ai-lab.github.io/hermit/docs/methods/ace.html)
+- [AHE Adapter](https://infini-ai-lab.github.io/hermit/docs/methods/ahe_adapter.html)
+- [RRSI Adapter](https://infini-ai-lab.github.io/hermit/docs/methods/rrsi_adapter.html)
+- [EvoSafeHarness](https://infini-ai-lab.github.io/hermit/docs/methods/evosafe.html)
 
 Hermit provides an RL-style training loop for evolving harness adapters:
 
@@ -77,12 +75,6 @@ Hermit provides an RL-style training loop for evolving harness adapters:
 <img src="assets/training-loop.gif" alt="Training pseudocode: initialize the adapter, run and judge tasks, call evolver.step, check and commit the proposal, then finalize." width="100%">
 </p>
 
-#### Run from the Command Line:
-
-<p align="center">
-<img src="assets/install-and-use.gif" alt="An animated command overview of installing Hermit, training an adapter, deploying it, and converting it into a plugin." width="100%">
-</p>
-
 ## Get Started
 
 ### Install
@@ -163,6 +155,8 @@ The generated `AGENTS.md` contains the learned rule. Start Codex in that workspa
 
 [Full walkthrough, expected outputs, and repeat-run instructions →](https://infini-ai-lab.github.io/hermit/docs/get-started/quickstart.html)
 
+**Run harness evolution with your coding plan.** Use your Codex/ChatGPT subscription for both acting and learning, with no separate model API key required. Plan-backed recipes are available for GEPA, SkillOpt, Harness², Continual Harness, ACE, AHE Adapter, and RRSI Adapter. See the [setup guide and recipes](https://infini-ai-lab.github.io/hermit/docs/recipes/officeqa.html#a-model-on-a-chatgpt-login) to get started; model access and usage follow your plan.
+
 **Ready for a real agent?** [Train with an API model](https://infini-ai-lab.github.io/hermit/docs/get-started/first-model.html) · [Run OfficeQA](https://infini-ai-lab.github.io/hermit/docs/recipes/officeqa.html) · [Bring your own task](https://infini-ai-lab.github.io/hermit/docs/tutorials/your-own-task.html)
 
 ### Examples and Recipes

diff --git a/docs/index.md b/docs/index.md
@@ -10,6 +10,7 @@ These docs describe the **0.2.0** code in this checkout.
 |---|---|---|
 | Run a complete training cycle without model access | [Quick Start](get-started/quickstart.md) | A changed adapter, held-out scores, and a host workspace |
 | Run with your model API | [First real-model training](get-started/first-model.md) | Acting calls, learner edits, and held-out evaluation on included data |
+| Run with your coding plan | [Harness evolution with your coding plan](tutorials/codex-plan.md) | A small benchmark run with Codex acting and learning |
 | Reproduce a benchmark experiment | [Recipe catalog](recipes/index.md) | A run using the recipe's data and model setup |
 | Choose a learning method | [Choose a method](methods/index.md) | A suitable template, Evolver, and split plan |
 | Improve your own agent | [Bring your task](tutorials/your-own-task.md) | A pipeline, judge, task description, and run config |
@@ -36,6 +37,7 @@ Passing file checks establishes structural validity. Performance improvement req
 get-started/installation
 get-started/quickstart
 get-started/first-model
+tutorials/codex-plan
 framework/what-is-a-harness-adapter
 faq
 ```

diff --git a/docs/tutorials/codex-plan.md 10000 b/docs/tutorials/codex-plan.md 10000
@@ -0,0 +1,191 @@
+# Run harness evolution with your coding plan
+
+Use your Codex / ChatGPT plan to run both the acting agent and the learner in Hermit. You can try methods such as
+GEPA, SkillOpt, ACE and RRSI Adapter without configuring a separate model API key or API billing account. Hermit
+runs the method's evaluation and selection loop; Codex supplies the model sessions that act and learn.
+
+This tutorial uses the existing **RRSI Adapter smoke run on OfficeQA**: four training questions, two trials per
+question, one round of two candidate adapters, then four held-out questions. It checks the integration rather than
+establishing a performance gain. It still needs the benchmark data and execution environment described below.
+
+## Start with a coding-agent prompt
+
+Open the Hermit checkout in your coding agent and give it this prompt. Replace the model placeholder with a model
+available to your Codex login.
+
+```text
+Help me run a small harness-evolution experiment using my Codex / ChatGPT plan.
+Use <my Codex model name> for both acting and learning, without a model API key.
+
+Read AGENTS.md, docs/tutorials/codex-plan.md and the RRSI Adapter example.
+Check the prerequisites, then use examples/officeqa/rrsi_adapter/smoke_luna.yaml.
+Keep its one-round protocol. Set acting concurrency to 2 and the training time
+budget to 2 hours. Use the same model settings for check, train and evaluation.
+
+Use the existing data-preparation scripts, container and egress-proxy setup.
+If the OfficeQA dataset or another prerequisite is missing, tell me exactly what
+I need to provide. Do not substitute synthetic data for the benchmark.
+
+Run the checks, train, and evaluate both the seed and final adapter. Inspect
+whether the round completed, which candidate was selected, and any errors.
+Report the commands, output paths and results. Keep generated artifacts local;
+do not commit or push anything.
+```
+
+The following sections explain the setup and commands behind that prompt.
+
+## 1. Prepare Codex and the benchmark
+
+Use a **Linux machine with Python 3.12+, Git, bubblewrap and Docker** for this example. RRSI's proposer and search
+roles use Hermit's Linux sandbox; acting sessions run in Docker. Complete [Installation](../get-started/installation.md)
+and install the two dependency groups from the repository root:
+
+```bash
+python -m pip install --upgrade "pip>=25.1"
+python -m pip install -e . --group rrsi --group skillopt
+```
+
+`rrsi` supplies the search method; `skillopt` supplies OfficeQA's shared data parsing and task utilities.
+
+Install [Codex CLI](https://learn.chatgpt.com/docs/codex/cli), then sign in with ChatGPT on the machine that will run
+Hermit. Hermit currently needs Codex's file-based login cache:
+
+```bash
+codex -c 'cli_auth_credentials_store="file"' login
+codex login status
+```
+
+Confirm that the authentication method is ChatGPT. Hermit reads `auth.json` under `CODEX_HOME`, or `~/.codex` when
+it is unset; a login held only in the OS credential store is not enough. Hermit manages per-session copies and
+refreshes through its shared login directory. See [OpenAI's authentication guide](https://learn.chatgpt.com/docs/auth)
+for sign-in and credential-storage options.
+
+Complete these existing OfficeQA setup steps once:
+
+- [Prepare the data](../recipes/officeqa.md#prepare-the-data): obtain the gated dataset, then run `prepare.py` with
+  the CSV and parsed corpus paths. This writes the task splits and initial skill used below. Dataset access is
+  separate from model authentication.
+- [Run in the Codex harness](../recipes/officeqa.md#run-in-the-codex-harness) and
+  [A model on a ChatGPT login](../recipes/officeqa.md#a-model-on-a-chatgpt-login): build the Linux Codex image,
+  create the internal Docker network, and start the login egress proxy. The example names
+  `officeqa-codex:0.164.0`, `oq-sandbox` and `oq-egress-login`. The image needs a Linux Codex distribution matching
+  Docker's architecture, even if you normally use Codex on another operating system.
+
+Then prepare RRSI's seed and trial rows:
+
+```bash
+python examples/officeqa/rrsi_adapter/make_seed.py
+python examples/officeqa/rrsi_adapter/make_rows.py
+```
+
+## 2. Check, train and evaluate
+
+Run from the repository root. Choose a model your login can actually use; the checked-in recipe's
+`gpt-5.6-luna` is the model used for its reported experiment, not a requirement to use your plan.
+
+```bash
+HERMIT_CODEX_MODEL='your-codex-model-name'
+
+hermit check --config examples/officeqa/rrsi_adapter/smoke_luna.yaml --no-probe \
+  agent.model.name="$HERMIT_CODEX_MODEL" evolver.model.name="$HERMIT_CODEX_MODEL" \
+  agent.concurrency=2 trainer.budget.hours=2
+
+hermit train --config examples/officeqa/rrsi_adapter/smoke_luna.yaml \
+  agent.model.name="$HERMIT_CODEX_MODEL" evolver.model.name="$HERMIT_CODEX_MODEL" \
+  agent.concurrency=2 trainer.budget.hours=2
+
+hermit eval --config examples/officeqa/rrsi_adapter/smoke_luna.yaml --adapter seed_adapter \
+  agent.model.name="$HERMIT_CODEX_MODEL" evolver.model.name="$HERMIT_CODEX_MODEL" \
+  agent.concurrency=2
+
+hermit eval --config examples/officeqa/rrsi_adapter/smoke_luna.yaml --adapter final \
+  agent.model.name="$HERMIT_CODEX_MODEL" evolver.model.name="$HERMIT_CODEX_MODEL" \
+  agent.concurrency=2
+```
+
+`check --no-probe` checks prerequisites without a task rollout; it does not prove model access or learner execution.
+Training normally evaluates H_0 and two candidates: **24 acting rollouts**, plus candidate liveness smokes and the
+proposer, analyst, digester and critic sessions. Each evaluation command adds four acting rollouts. A small task
+split can still consume substantial plan usage. The two-hour training budget can end the run before its round
+finishes; probes and separate evaluations are outside that training budget.
+
+Outputs are under `examples/officeqa/rrsi_adapter/runs/smoke-luna-rrsi/`. Inspect:
+
+- `checkout/ADAPTER.md` and any additional learned files;
+- `step_NNNN/output/report.md` for candidate screening, round decisions and the selected incumbent;
+- `workspace/state/rrsi/frontier.json` for the incumbent and round trajectory;
+- `eval/seed_adapter/summary.json` and `eval/final/summary.json` for held-out scores and execution errors.
+
+Check that the round completed before treating the output as a completed RRSI smoke. `final` is the incumbent H_T;
+an interrupted run can return its last incumbent. A run that completes with no accepted candidate is also valid.
+Keep model and task settings fixed when comparing seed and final scores.
+
+For a separate experiment, change both `trainer.run_name` and `trainer.tree_dir`; apply both overrides consistently
+to its training and evaluation commands. An unfinished run resumes automatically. Keep generated records and adapter snapshots local,
+as described in [Contributing](https://github.com/Infini-AI-Lab/hermit/blob/main/CONTRIBUTING.md).
+
+## How this works for other methods
+
+There are two learner configurations, depending on how the method calls its model:
+
+| Method's learner | Configuration | Examples |
+|---|---|---|
+| Requests text replies for reflection or revision | `llm.cli: codex` with `login: chatgpt`; each request becomes a Codex session in its own container | GEPA, SkillOpt, ACE, Continual Harness |
+| Edits files as an agent | `evolver.base: codex` and `evolver.model.login: chatgpt`; sessions run in the method's workspace | Harness², Meta-Harness Adapter, AHE Adapter, RRSI Adapter |
+
+For the first form, the model settings look like this; use the method's existing recipe for its remaining fields:
+
+```yaml
+llm:
+  cli: codex
+  model: your-codex-model-name
+  login: chatgpt
+  effort: medium
+  web_search: false
+  isolation: container
+  container:
+    image: officeqa-codex:0.164.0
+    network: oq-sandbox
+    proxy: http://oq-egress-login:3128
+  timeout: 3600
+```
+
+For the second form, keep the method's `evolver.name` and parameters, and configure its Base Agent and model:
+
+```yaml
+evolver:
+  name: rrsi_adapter
+  # Keep the recipe's method-specific params here.
+  base: codex
+  model: {name: your-codex-model-name, login: chatgpt, effort: medium}
+```
+
+The acting agent also needs `agent.base: codex` and `agent.model.login: chatgpt`, as in the smoke configuration.
+Changing only the learner does not change an API-based actor. Use a complete recipe for the task's preparation,
+scoring, container and sandbox settings. For example, [ACE](../methods/ace.md) already provides
+`examples/officeqa/ace/smoke_codex_luna.yaml`; GEPA and SkillOpt have `run_codex_luna.yaml` recipes.
+
+Plan-backed sessions use the account's available models and usage limits. Hermit reports unknown dollar cost when
+there is no per-token price; use time/token budgets and inspect the shared plan status described in
+[the login setup](../recipes/officeqa.md#a-model-on-a-chatgpt-login). Extra services still need their own setup:
+ReHarness's embedding model, for example, is not provided by this Codex text-model route.
+
+## What about a Claude Code plan?
+
+Claude Code also supports subscription sign-in, including Pro and Max; see
+[Claude's subscription setup](https://support.claude.com/en/articles/11145838-use-claude-code-with-your-pro-max-or-ultra-plan).
+Hermit already has a Claude Code Base Agent, but its current coverage differs from the Codex walkthrough above:
+
+| Path | Current support |
+|---|---|
+| Acting agent | `agent.base: claude` uses Claude Code with the task's preparation and scoring functions. The current container runner is Codex-only, so it needs its own execution setup. |
+| Agent-style learner | The built-in `claude` Evolver and Meta-Harness Adapter support Claude Code sessions. RRSI Adapter supports a Claude proposer, but its analyst, digesters and critic still require Codex. AHE Adapter's learners are Codex-only. |
+| Text-reply learner (`llm.cli`) | Only `codex` is accepted today. Switching GEPA, SkillOpt or ACE to `llm.cli: claude` requires shared-layer support first. |
+
+Hermit's Claude sign-in path copies `~/.claude/.credentials.json` into the session's home when no API binding is
+configured; it does not read a login stored only in macOS Keychain. The Codex-specific `login: chatgpt`, container
+and plan-management settings above do not configure Claude Code. A complete Claude-plan recipe has not been
+validated in this tutorial.
+
+Continue with [Choose a method](../methods/index.md), the [OfficeQA recipe](../recipes/officeqa.md), or
+[Bring your task](your-own-task.md) to use your own benchmark.