@litfamily/litopencode 1.0.4 → 1.0.5

package/README.md

@@ -1,76 +1,59 @@
-<p align="center"><picture><source media="(prefers-reduced-motion: reduce)" srcset="https://cdn.jsdelivr.net/npm/@litfamily/litopencode@1.0.4/docs/assets/cover.webp" /><img src="https://cdn.jsdelivr.net/npm/@litfamily/litopencode@1.0.4/docs/assets/cover-motion.webp" width="100%" alt="LitOpenCode cover art. A lime signal moves through orbital paths above the OPENCODE wordmark. The static family cover appears for reduced-motion preference." /></picture></p>
-<p align="center"><a href="https://cdn.jsdelivr.net/npm/@litfamily/litopencode@1.0.4/docs/assets/cover.webp">View the static cover</a></p>
+<p align="center"><picture><source media="(prefers-reduced-motion: reduce)" srcset="https://cdn.jsdelivr.net/npm/@litfamily/litopencode@1.0.5/docs/assets/cover-motion-still.webp" /><img src="https://cdn.jsdelivr.net/npm/@litfamily/litopencode@1.0.5/docs/assets/cover-motion.webp" width="100%" alt="LitOpenCode cover art. A lime signal traces orbital paths above the OPENCODE wordmark. Reduced-motion users see a still frame from this cover." /></picture></p>
+<p align="center"><a href="https://cdn.jsdelivr.net/npm/@litfamily/litopencode@1.0.5/docs/assets/cover.webp">View the static cover</a></p>
 
-# LitOpenCode
-
-[English](README.md) · [한국어](README-Ko-KR.md)
-
-**Keep the work lit.**
-
-<p align="center"><img src="https://cdn.jsdelivr.net/npm/@litfamily/litopencode@1.0.4/docs/assets/readme/litopencode-wordmark.svg" width="480" alt="LITOPENCODE display type" /></p>
-<p align="center"><img src="https://cdn.jsdelivr.net/npm/@litfamily/litopencode@1.0.4/docs/assets/readme/litopencode-clay-icon.png" width="160" alt="LitOpenCode clay mark" /></p>
+<p align="center"><img src="https://cdn.jsdelivr.net/npm/@litfamily/litopencode@1.0.5/docs/assets/readme/ascii-readme.svg" width="480" alt="LIT ASCII B mark" /></p>
 
-> **The spark is yours.**<br>
-> **Bring it to your work.**
-
-A bug to fix. A screen to build. A project to finish.
-
-Starting takes a line. Continuing takes more. When a conversation grows long or a
-session ends, you have to recover where you stopped: what you decided, what you
-checked, and what comes next.
-
-**LIT keeps that spark in the project.** It records the goal, the plan, verified
-results, and the next steps so another session can read them and continue.
-
-**Where one conversation ends, the next piece of work begins.**
-
-LitOpenCode adds workflow agents, slash commands, and a local evidence ledger to
-**OpenCode**. Keep the tool you already use; no other LitFamily product is required.
-A lasting flame does not mean a program runs forever. **It means leaving work ready
-to continue when the session ends.**
+<details>
+<summary>Copy ASCII logo</summary>
 
 ```text
-Plan → Build → Verify → Leave the next step
-```
-
-[Install](#install-in-30-seconds) · [First use](#first-use) · [Core routes](#core-routes) · [Safety](#safety-and-updates) · [Docs](#documentation)
-
-## Start with one lit task
-
-Append `lit` to a prompt, or use `/lit`, to select the LitOpenCode workflow. OpenCode owns the model and permission prompts; the plugin keeps project records.
-
-| Prompt or route | Effect |
-| --- | --- |
-| `lit` or `/lit` | Start a bounded task and record what was checked. |
-| `handoff` or `/lit-handoff` | Carry the current result and next step into another session. |
-| `lit-plan` | Prepare a plan before implementation. |
-| `/start-work <approved-plan>` | Execute an approved plan. |
-| `/review-work` | Review the change and evidence. |
-| `/litresearch` | Use the shipped source-backed research route. |
-
-Selecting a route does not claim that the model ran or that a visual check passed.
+                             ▄▄▄▄
+                   ▗███▌   ▗██████▖
+ ▗▄▄▄▄▄          ▗▟████▌   ▝██████▘
+ ▐█████        ▗▟██████▌    ▝▀▜█▀▘
+ ▐█████      ▗▟███████▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
+ ▐█████    ▗▟█████████████████████████ ▐█▀
+ ▐█████    ████████████████████████████▀
+ ▐█████    ██▛▘   ▄ ▄▄▄▄▖▄▄▄▄▄▄▄▄▄▄▄▄▄▖
+ ▐█████    ▀    ▄██ ████▌█████████████▌
+ ▐█████       ▄████ ████▌█████████████▌
+ ▐█████     ▄█████▛                                 litopencode
+ ▐█████  ▗▟█████▀▘       ▄▄▄▄▄     ▗▖
+ ▐█████ ▐█████▀          █████     ▐▛▀
+ ▐█████ ▐███▀            █████
+ ▐█████ ▐█▀              █████
+ ▐█████ ▝                █████
+ ▐█████▄▄▄▄▄▄▄▖          █████
+ ▐███████████▛           █████
+ ▐██████████▀            █████
 
-The packaged reference corpora use the canonical paths `vendor/handoff/` and
-`vendor/scientific-visualization/`. Their numbered license and provenance filenames
-remain source-attribution records.
+```
 
-## Practical routes and visual cues
+</details>
 
-Use `lit` or `/lit` to start a bounded task; `handoff` or `/lit-handoff` carries the checked result and next step; `lit-plan` prepares a plan; `/start-work <approved-plan>` executes an approved plan; `/review-work` checks the change and its evidence; and `/litresearch` uses the shipped source-backed research route. The expected effect is a project-local goal, evidence, and next-step record that another session can resume. The host limit is explicit: OpenCode owns model execution and permission prompts, while the plugin supplies routes and records; selecting a route does not prove execution or a visual check.
+# LitOpenCode
 
-When a Lit route activates, the reply starts with a bold ignition line. The plugin requests a six-second warning toast with five micro-logo rows and a final `🔥 LIT IGNITED · <discipline> 🔥` line when the environment supports mark glyphs; otherwise, it shows only that line.
+[English](https://cdn.jsdelivr.net/npm/@litfamily/litopencode@1.0.5/README.md) · [한국어](https://cdn.jsdelivr.net/npm/@litfamily/litopencode@1.0.5/README-Ko-KR.md)
 
-<p align="center"><img src="https://cdn.jsdelivr.net/npm/@litfamily/litopencode@1.0.4/docs/assets/litopencode-ignition-1600.webp" width="48%" alt="LitOpenCode ignition apparatus" /><img src="https://cdn.jsdelivr.net/npm/@litfamily/litopencode@1.0.4/docs/assets/litopencode-continuity-1600.webp" width="48%" alt="LitOpenCode continuity apparatus" /></p>
+**Keep the work lit.**
 
-<p align="center"><a href="docs/assets/readme/ignition-film.mp4"><img src="docs/assets/readme/poster.png" width="720" alt="Ignition motion poster" /></a></p>
+<p align="center"><img src="https://cdn.jsdelivr.net/npm/@litfamily/litopencode@1.0.5/docs/assets/readme/litopencode-wordmark.svg" width="480" alt="LITOPENCODE display type" /></p>
+<p align="center"><img src="https://cdn.jsdelivr.net/npm/@litfamily/litopencode@1.0.5/docs/assets/readme/litopencode-clay-icon.png" width="160" alt="LitOpenCode clay mark" /></p>
 
-The poster opens the optional film; this README keeps motion opt-in.
+<p align="center">
+<a href="#install"><img src="https://cdn.jsdelivr.net/npm/@litfamily/litopencode@1.0.5/docs/assets/readme/badge-version.svg" alt="1.0.5" /></a>
+<a href="https://cdn.jsdelivr.net/npm/@litfamily/litopencode@1.0.5/LICENSE"><img src="https://cdn.jsdelivr.net/npm/@litfamily/litopencode@1.0.5/docs/assets/readme/badge-license.svg" alt="MIT license" /></a>
+</p>
 
-The retained `docs/assets/cover.svg` is the editable, low-bandwidth fallback; the WebP above is the product emphasis export.
+<p align="center">
+<a href="https://cdn.jsdelivr.net/npm/@litfamily/litopencode@1.0.5/docs/reference.md"><img src="https://cdn.jsdelivr.net/npm/@litfamily/litopencode@1.0.5/docs/assets/readme/lucide-book-open.svg" width="16" alt="" /> Docs</a> &nbsp; <a href="#install">Install</a> &nbsp; <a href="#ignition-motion"><img src="https://cdn.jsdelivr.net/npm/@litfamily/litopencode@1.0.5/docs/assets/readme/lucide-play.svg" width="16" alt="" /> Ignition</a> &nbsp; <a href="https://cdn.jsdelivr.net/npm/@litfamily/litopencode@1.0.5/LICENSE"><img src="https://cdn.jsdelivr.net/npm/@litfamily/litopencode@1.0.5/docs/assets/readme/lucide-shield-check.svg" width="16" alt="" /> MIT</a>
+</p>
 
+## What is LitOpenCode
 
+LitOpenCode adds workflow agents, slash commands, and a local evidence ledger to OpenCode.
 
-## Install in 30 seconds
+## Install
 
 Install the scoped package with Node.js/npm and OpenCode available:
 
@@ -83,19 +66,20 @@ Restart OpenCode, press **Tab**, and select **lit-loop**. Model access and crede
 come from your OpenCode provider setup. The installer offers model, permission, and
 output-style choices; the default permission mode is safe/ask-first.
 
+With the OpenAI provider, fresh installs use GPT-6 Astra (`gpt-6-astra`) at `xhigh` for planning and review, and GPT-6 Luna (`gpt-6-luna`) at `max` for execution and research. The installer also offers supported model and effort choices.
+
 ```sh
 npm exec --package @litfamily/litopencode@latest -- litopencode install --dry-run  # preview changes
 npm exec --package @litfamily/litopencode@latest -- litopencode install --yes     # use defaults; preserve saved choices
-npm exec --package @litfamily/litopencode@latest -- litopencode doctor            # inspect package and configuration health
 ```
 
 Installation registers the plugin and native command/skill files under your OpenCode
 config root. Routes live in `~/.config/opencode/litopencode.json`; `XDG_CONFIG_HOME`
 changes that root. Existing custom routes remain user-owned. See the
-[installation and model reference](docs/reference.md#install) for custom roots,
+[installation and model reference](https://cdn.jsdelivr.net/npm/@litfamily/litopencode@1.0.5/docs/reference.md#install) for custom roots,
 provider choices, terminal policy, and unattended setup.
 
-This checkout is `@litfamily/litopencode@1.0.4`. Registry `@latest` can differ; inspect it with
+This checkout is `@litfamily/litopencode@1.0.5`. Registry `@latest` can differ; inspect it with
 `npm view @litfamily/litopencode version` when an exact version matters.
 
 ## First use
@@ -108,7 +92,7 @@ lit Build a to-do list in one index.html with no dependencies; check adding and
 ```
 
 Your first result is `index.html`. Open it and try adding and completing an item.
-If the behavior differs from the completion report, tell the agent what happened.
+
 When you return, use `/lit-recap` to read the record and find the next step.
 
 Use `/lit` for the same workflow. For work that needs a reviewed plan, select
@@ -120,48 +104,45 @@ Progress and evidence live in `.litopencode/litgoal/`. OpenCode does not expose
 native goal primitive; this repository-local ledger supports resuming work. Static
 skills provide guidance when their documented routes select them.
 
-## Ignition mark
+## Key features
 
-<p align="center"><img src="docs/assets/readme/ascii-readme.svg" width="480" alt="LIT ASCII B mark" /></p>
+> **The spark is yours.**<br>
+> **Bring it to your work.**
 
-<details>
-<summary>Copy ASCII logo</summary>
+A bug to fix. A screen to build. A project to finish.
 
-```text
-                             ▄▄▄▄
-                   ▗███▌   ▗██████▖
- ▗▄▄▄▄▄          ▗▟████▌   ▝██████▘
- ▐█████        ▗▟██████▌    ▝▀▜█▀▘
- ▐█████      ▗▟███████▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
- ▐█████    ▗▟█████████████████████████ ▐█▀
- ▐█████    ████████████████████████████▀
- ▐█████    ██▛▘   ▄ ▄▄▄▄▖▄▄▄▄▄▄▄▄▄▄▄▄▄▖
- ▐█████    ▀    ▄██ ████▌█████████████▌
- ▐█████       ▄████ ████▌█████████████▌
- ▐█████     ▄█████▛                                 litopencode
- ▐█████  ▗▟█████▀▘       ▄▄▄▄▄     ▗▖
- ▐█████ ▐█████▀          █████     ▐▛▀
- ▐█████ ▐███▀            █████
- ▐█████ ▐█▀              █████
- ▐█████ ▝                █████
- ▐█████▄▄▄▄▄▄▄▖          █████
- ▐███████████▛           █████
- ▐██████████▀            █████
+Starting takes a line. Continuing takes more. When a conversation grows long or a
+session ends, you have to recover where you stopped: what you decided, what you
+checked, and what comes next.
+
+**LIT keeps that spark in the project.** It records the goal, the plan, verified
+results, and the next steps so another session can read them and continue.
+
+**Where one conversation ends, the next piece of work begins.**
 
+ Keep the tool you already use; no other LitFamily product is required.
+A lasting flame does not mean a program runs forever. **It means leaving work ready
+to continue when the session ends.**
+
+```text
+Plan → Build → Verify → Leave the next step
 ```
 
-</details>
+### Practical routes and visual cues
 
-<p align="center">
-<a href="#install-in-30-seconds"><img src="https://cdn.jsdelivr.net/npm/@litfamily/litopencode@1.0.4/docs/assets/readme/badge-version.svg" alt="1.0.4" /></a>
-<a href="LICENSE"><img src="https://cdn.jsdelivr.net/npm/@litfamily/litopencode@1.0.4/docs/assets/readme/badge-license.svg" alt="MIT license" /></a>
-</p>
+Use `lit` or `/lit` to start a bounded task; `handoff` or `/lit-handoff` carries the checked result and next step; `lit-plan` prepares a plan; `/start-work <approved-plan>` executes an approved plan; `/review-work` checks the change and its evidence; and `/litresearch` uses the shipped source-backed research route. The expected effect is a project-local goal, evidence, and next-step record that another session can resume. The host limit is explicit: OpenCode owns model execution and permission prompts, while the plugin supplies routes and records; selecting a route does not prove execution or a visual check.
 
-<p align="center">
-<a href="docs/reference.md"><img src="docs/assets/readme/lucide-book-open.svg" width="16" alt="" /> Docs</a> &nbsp; <a href="#install-in-30-seconds">Install</a> &nbsp; <a href="#ignition-motion"><img src="docs/assets/readme/lucide-play.svg" width="16" alt="" /> Ignition</a> &nbsp; <a href="LICENSE"><img src="docs/assets/readme/lucide-shield-check.svg" width="16" alt="" /> MIT</a>
-</p>
+When a Lit route activates, the reply starts with a bold ignition line. The plugin requests a six-second warning toast with five micro-logo rows and a final `🔥 LIT IGNITED · <discipline> 🔥` line when the environment supports mark glyphs; otherwise, it shows only that line.
+
+<p align="center"><img src="https://cdn.jsdelivr.net/npm/@litfamily/litopencode@1.0.5/docs/assets/litopencode-ignition-1600.webp" width="48%" alt="LitOpenCode ignition apparatus" /><img src="https://cdn.jsdelivr.net/npm/@litfamily/litopencode@1.0.5/docs/assets/litopencode-continuity-1600.webp" width="48%" alt="LitOpenCode continuity apparatus" /></p>
 
-## How LitOpenCode fits into OpenCode
+<p align="center"><a href="https://cdn.jsdelivr.net/npm/@litfamily/litopencode@1.0.5/docs/assets/readme/ignition-film.mp4"><img src="https://cdn.jsdelivr.net/npm/@litfamily/litopencode@1.0.5/docs/assets/readme/poster.png" width="720" alt="Ignition motion poster" /></a></p>
+
+The poster opens the optional film; this README keeps motion opt-in.
+
+The retained `docs/assets/cover.svg` is the editable, low-bandwidth fallback; the WebP above is the product emphasis export.
+
+### How LitOpenCode fits into OpenCode
 
 OpenCode loads the plugin from its config. Agents and chat/command routes select
 workflows and installed skills. Tools and lifecycle hooks keep the project records
@@ -185,14 +166,34 @@ flowchart TD
     class Skills,Records local
 ```
 
-## LITFAMILY
+### Design production and README Studio
 
-![Five armored machines representing LitClaude, LitHermes, LitCodex, LitOpenCode and LitGrok](docs/assets/readme/litfamily-machines.png)
+The native `frontend-ui-ux` skill takes an adequate authorized build request through working implementation and rendered inspection. It asks only material design choices and retains answers; review and plan requests remain read-only.
 
-LitClaude · LitHermes · LitCodex · LitOpenCode · LitGrok.
-Five armored machines for five independent products, each working in its own host.
+Select `readme-studio` through OpenCode's native skill tool for factual README writing and local cover composition. The installed resources include outlined typography helpers and pinned Remotion/HyperFrames recipes. Without a native image generator it reports `IMAGE_GENERATION_UNAVAILABLE` and can continue from a supplied background. Public GitHub/npm rendering is a separate gate. No dedicated slash command is claimed.
+
+## Commands
 
-## Core routes
+### Start with one lit task
+
+Append `lit` to a prompt, or use `/lit`, to select the LitOpenCode workflow. OpenCode owns the model and permission prompts; the plugin keeps project records.
+
+| Prompt or route | Effect |
+| --- | --- |
+| `lit` or `/lit` | Start a bounded task and record what was checked. |
+| `handoff` or `/lit-handoff` | Carry the current result and next step into another session. |
+| `lit-plan` | Prepare a plan before implementation. |
+| `/start-work <approved-plan>` | Execute an approved plan. |
+| `/review-work` | Review the change and evidence. |
+| `/litresearch` | Use the shipped source-backed research route. |
+
+Selecting a route does not claim that the model ran or that a visual check passed.
+
+The packaged reference corpora use the canonical paths `vendor/handoff/` and
+`vendor/scientific-visualization/`. Their numbered license and provenance filenames
+remain source-attribution records.
+
+### Core routes
 
 | Route | Use it to |
 | --- | --- |
@@ -205,9 +206,9 @@ Five armored machines for five independent products, each working in its own hos
 | `/lit-korean` | Improve Korean prose without changing its meaning. |
 | `/lit-scientific-visualization` | Use the packaged scientific-visualization workflow. |
 
-The [full route and skill reference](docs/reference.md#core-commands) covers
+The [full route and skill reference](https://cdn.jsdelivr.net/npm/@litfamily/litopencode@1.0.5/docs/reference.md#core-commands) covers
 Autoresearch, Autoconference, Wikify, UI/UX, and the two-lane repository rules engine.
-[Migration notes](docs/migration.md#skill-id-renames) explain the one-release skill aliases.
+[Migration notes](https://cdn.jsdelivr.net/npm/@litfamily/litopencode@1.0.5/docs/migration.md#skill-id-renames) explain the one-release skill aliases.
 
 ## Safety and updates
 
@@ -218,7 +219,19 @@ Autoresearch, Autoconference, Wikify, UI/UX, and the two-lane repository rules e
   check. Use `--no-auto-update` or `LITOPENCODE_NO_AUTO_UPDATE=1` to disable automatic updates.
 - Skill learning requires an explicit apply command. Its review, mutation, rollback, and
   curator boundaries support POSIX; on Windows those boundaries fail closed. Install,
-  doctor, and passive surfaces remain available. [Learning-loop details](docs/reference.md#skill-learning-loop).
+  doctor, and passive surfaces remain available. [Learning-loop details](https://cdn.jsdelivr.net/npm/@litfamily/litopencode@1.0.5/docs/reference.md#skill-learning-loop).
+
+## Troubleshooting
+
+If the behavior differs from the completion report, tell the agent what happened.
+
+Inspect package and configuration health with:
+
+```sh
+npm exec --package @litfamily/litopencode@latest -- litopencode doctor
+```
+
+See the [installation and model reference](https://cdn.jsdelivr.net/npm/@litfamily/litopencode@1.0.5/docs/reference.md#install).
 
 ## Uninstall
 
@@ -234,22 +247,37 @@ npm uninstall -g @litfamily/litopencode
 Review installed command/skill files before removing installer-owned copies; preserve
 user-created files and edits. Keep `litopencode.json` if you still need its routes, and
 keep project `.litopencode/` ledgers if you plan to resume. Restart OpenCode after removal.
-See the [removal reference](docs/reference.md#remove).
+See the [removal reference](https://cdn.jsdelivr.net/npm/@litfamily/litopencode@1.0.5/docs/reference.md#remove).
+
+## License
+
+MIT
+
+## Links
+
+### Documentation
 
-## Documentation
+- [Workflow reference](https://cdn.jsdelivr.net/npm/@litfamily/litopencode@1.0.5/docs/reference.md): models, permissions, host hooks, all skills, learning loop, and reproducible verification.
+- [한국어 상세 안내](https://cdn.jsdelivr.net/npm/@litfamily/litopencode@1.0.5/docs/reference-Ko-KR.md)
+- [Migration notes](https://cdn.jsdelivr.net/npm/@litfamily/litopencode@1.0.5/docs/migration.md)
+- [Terminal mark and activation probes](https://cdn.jsdelivr.net/npm/@litfamily/litopencode@1.0.5/docs/lit-mark.md)
+- [Changelog](https://cdn.jsdelivr.net/npm/@litfamily/litopencode@1.0.5/CHANGELOG.md)
+- Maintainer release checklist (repository only)
 
-- [Workflow reference](docs/reference.md): models, permissions, host hooks, all skills, learning loop, and reproducible verification.
-- [한국어 상세 안내](docs/reference-Ko-KR.md)
-- [Migration notes](docs/migration.md)
-- [Terminal mark and activation probes](docs/lit-mark.md)
-- [Changelog](CHANGELOG.md)
-- [Maintainer release checklist](docs/release-checklist.md) (repository only)
+- [Contributing](https://cdn.jsdelivr.net/npm/@litfamily/litopencode@1.0.5/CONTRIBUTING.md) · [Support](https://cdn.jsdelivr.net/npm/@litfamily/litopencode@1.0.5/SUPPORT.md) · [Security](https://cdn.jsdelivr.net/npm/@litfamily/litopencode@1.0.5/SECURITY.md)
+- [Code of conduct](https://cdn.jsdelivr.net/npm/@litfamily/litopencode@1.0.5/CODE_OF_CONDUCT.md) · [Privacy and network behavior](https://cdn.jsdelivr.net/npm/@litfamily/litopencode@1.0.5/docs/privacy.md)
+
+### LITFAMILY
+
+![Five armored machines representing LitClaude, LitHermes, LitCodex, LitOpenCode and LitGrok](https://cdn.jsdelivr.net/npm/@litfamily/litopencode@1.0.5/docs/assets/readme/litfamily-machines.png)
+
+LitClaude · LitHermes · LitCodex · LitOpenCode · LitGrok.
+Five armored machines for five independent products, each working in its own host.
 
-- [Contributing](CONTRIBUTING.md) · [Support](SUPPORT.md) · [Security](SECURITY.md)
-- [Code of conduct](CODE_OF_CONDUCT.md) · [Privacy and network behavior](docs/privacy.md)
+![Five robots beneath the LITFAMILY and KEEP THE WORK LIT. lettering.](https://cdn.jsdelivr.net/npm/@litfamily/litopencode@1.0.5/docs/assets/cover.webp)
 
-## Ignition motion
+### Ignition motion
 
-[![Ignition motion graphic poster](docs/assets/readme/poster.png)](docs/assets/readme/ignition-film.mp4)
+[![Ignition motion graphic poster](https://cdn.jsdelivr.net/npm/@litfamily/litopencode@1.0.5/docs/assets/readme/poster.png)](https://cdn.jsdelivr.net/npm/@litfamily/litopencode@1.0.5/docs/assets/readme/ignition-film.mp4)
 
-[Watch the 10-second film](docs/assets/readme/ignition-film.mp4) · [Animated GIF](docs/assets/readme/ignition-readme.gif) · [Lucide icon license](docs/assets/readme/Lucide-LICENSE.txt) · [ASCII font license](docs/assets/readme/JetBrainsMono-OFL.txt)
+[Watch the 10-second film](https://cdn.jsdelivr.net/npm/@litfamily/litopencode@1.0.5/docs/assets/readme/ignition-film.mp4) · [Animated GIF](https://cdn.jsdelivr.net/npm/@litfamily/litopencode@1.0.5/docs/assets/readme/ignition-readme.gif) · [Lucide icon license](https://cdn.jsdelivr.net/npm/@litfamily/litopencode@1.0.5/docs/assets/readme/Lucide-LICENSE.txt) · [ASCII font license](https://cdn.jsdelivr.net/npm/@litfamily/litopencode@1.0.5/docs/assets/readme/JetBrainsMono-OFL.txt)

package/dist/activation-workflow-prompts.js

@@ -174,7 +174,7 @@ Stop when nothing left is unknown that would change the plan. Over-interviewing
 Hand off a brief that states the objective in product terms, explicit non-goals, the decisions taken and who made each, the assumptions with their fallbacks, the acceptance criteria, and the unknowns deliberately left open. Then hand it to lit-plan; do not plan here.
 `;
 export const deepInterviewPromptInjection = skillBackedPromptInjection("deep-interview-mode", deepInterviewActivationGuidance, "deep-interview");
-export const frontendUiUxPromptInjection = skillBackedPromptInjection("frontend-ui-ux-mode", "", "frontend-ui-ux");
+export const frontendUiUxPromptInjection = skillBackedPromptInjection("frontend-ui-ux-mode", "Honor build, review and plan intent. Continue an adequately specified authorized build through implementation and inspection.", "frontend-ui-ux");
 const browserDriveActivationGuidance = `
 Activate browser-drive. The answer lives in a running page, and the first action is to find out whether a driver exists here.
 

package/dist/agents/registry.js

@@ -2,20 +2,19 @@ import { litOpenCodeDefaultAgents } from "./defaults.js";
 import { litOpenCodeSpecialistAgents } from "./specialists.js";
 import { withReaderFacingCommunicationContract } from "../reader-facing-output.js";
 import { defaultGlobalConfigFile, configSectionWasAuthored, effectiveAgentConfig } from "../config.js";
+import { defaultOpenAiRoutes, newGenerationOpenAiModels, openAiModelEfforts, openaiProviderModels } from "../cli/model-catalog.js";
 import { assertSafeEffectiveAgentRoutes } from "../model-route-policy.js";
 export const litOpenCodeAgents = Object.freeze([
     ...litOpenCodeDefaultAgents,
     ...litOpenCodeSpecialistAgents
 ]);
-const hostLeadRoute = Object.freeze({ model: "openai/gpt-6-astra", variant: "xhigh", reasoningEffort: "xhigh" });
-const specialLeadAgentIds = Object.freeze(["momus", "litwork-reviewer"]);
-const astraEfforts = new Set(["low", "medium", "high", "xhigh", "max"]);
-const astraModelMetadata = Object.freeze({
-    name: "GPT-6 Astra",
-    reasoning: true,
-    temperature: false,
-    tool_call: true
+const hostLeadRoute = Object.freeze({
+    model: "openai/" + defaultOpenAiRoutes.lead.model,
+    variant: defaultOpenAiRoutes.lead.effort,
+    reasoningEffort: defaultOpenAiRoutes.lead.effort
 });
+const specialLeadAgentIds = Object.freeze(["momus", "litwork-reviewer"]);
+const astraEfforts = new Set(openAiModelEfforts(newGenerationOpenAiModels.astra));
 function toolsToConfig(tools) {
     const enabledTools = {};
     for (const tool of tools) {
@@ -201,30 +200,35 @@ function applySpecialLeadRoute(agentTarget, agentId) {
         return;
     agentTarget[agentId] = { ...existing, ...hostLeadRoute };
 }
-function registerAstraProviderModel(config) {
+function registerOpenAiProviderModels(config) {
     const providers = config.provider ?? {};
     const openai = providers.openai;
     const models = openai?.models ?? {};
-    const existingAstra = models["gpt-6-astra"];
-    const astraModel = existingAstra === undefined
-        ? { ...astraModelMetadata }
-        : {
-            ...existingAstra,
-            ...(existingAstra.name === undefined ? { name: astraModelMetadata.name } : {}),
-            ...(existingAstra.reasoning === undefined ? { reasoning: astraModelMetadata.reasoning } : {}),
-            ...(existingAstra.temperature === undefined ? { temperature: astraModelMetadata.temperature } : {}),
-            ...(existingAstra.tool_call === undefined ? { tool_call: astraModelMetadata.tool_call } : {})
-        };
-    if (existingAstra !== undefined && JSON.stringify(existingAstra) === JSON.stringify(astraModel))
+    const nextModels = { ...models };
+    let changed = false;
+    for (const [modelId, metadata] of Object.entries(openaiProviderModels)) {
+        const existing = models[modelId];
+        const next = existing === undefined
+            ? { ...metadata }
+            : {
+                ...existing,
+                ...(existing.name === undefined ? { name: metadata.name } : {}),
+                ...(existing.reasoning === undefined ? { reasoning: metadata.reasoning } : {}),
+                ...(existing.temperature === undefined ? { temperature: metadata.temperature } : {}),
+                ...(existing.tool_call === undefined ? { tool_call: metadata.tool_call } : {})
+            };
+        if (JSON.stringify(existing) !== JSON.stringify(next)) {
+            nextModels[modelId] = next;
+            changed = true;
+        }
+    }
+    if (!changed)
         return;
     config.provider = {
         ...providers,
         openai: {
             ...(openai ?? {}),
-            models: {
-                ...models,
-                "gpt-6-astra": astraModel
-            }
+            models: nextModels
         }
     };
 }
@@ -234,7 +238,7 @@ function isAstraEffort(value) {
 function modelFields(override) {
     const model = openCodeModel(override);
     const nativeReasoningEffort = override.reasoningEffort ??
-        (model === "openai/gpt-6-astra" && override.variant !== undefined && isAstraEffort(override.variant)
+        (model === "openai/" + newGenerationOpenAiModels.astra && override.variant !== undefined && isAstraEffort(override.variant)
             ? override.variant
             : undefined);
     return {
@@ -270,7 +274,7 @@ export function registerLitOpenCodeAgents(config, litConfig = defaultGlobalConfi
         ...specialLeadAgentIds.filter((agentId) => litConfig.agents[agentId] !== undefined)
     ];
     assertSafeEffectiveAgentRoutes(litConfig, routeAgentIds);
-    registerAstraProviderModel(config);
+    registerOpenAiProviderModels(config);
     const existingBuild = config.agent?.build ?? {};
     const existingPlan = config.agent?.plan ?? {};
     const inheritedTaskPermission = taskPermission(config.permission);

package/dist/cli/args.js

@@ -1,6 +1,6 @@
 import os from "node:os";
 import path from "node:path";
-import { resolveModelRoute } from "./model-catalog.js";
+import { installModelEfforts, isInstallModelEffort, resolveModelRoute } from "./model-catalog.js";
 export const loopCommands = [
     "create-goals",
     "status",
@@ -332,9 +332,9 @@ function parsePermissionMode(value, optionName) {
 }
 function parseModelEffort(value, optionName) {
     const cleaned = cleanOption(value, optionName);
-    if (cleaned === "low" || cleaned === "medium" || cleaned === "high" || cleaned === "xhigh" || cleaned === "max")
+    if (cleaned !== undefined && isInstallModelEffort(cleaned))
         return cleaned;
-    throw new Error(optionName + " must be one of low, medium, high, xhigh, max");
+    throw new Error(optionName + " must be one of " + installModelEfforts.join(", "));
 }
 function cleanOption(value, optionName) {
     if (value === undefined)

package/dist/cli/host-limits.d.ts

@@ -1,7 +1,7 @@
 export declare const contextCeilingTokens = 372000;
 export declare const compactionTriggerTokens = 334800;
 export declare const compactionReserveTokens: number;
-export declare const offeredGpt56Models: readonly ["gpt-5.6-luna"];
+export declare const offeredOpenAiModels: readonly ["gpt-6-luna", "gpt-5.6-luna"];
 export type HostLimitMutation = {
     readonly reconfigure: boolean;
     readonly changed: boolean;

package/dist/cli/host-limits.js

@@ -1,11 +1,12 @@
 import { isRecord } from "./json.js";
 import { writeTextAtomically } from "./install-config.js";
+import { hostLimitModelIds } from "./model-catalog.js";
 import fs from "node:fs/promises";
 import path from "node:path";
 export const contextCeilingTokens = 372000;
 export const compactionTriggerTokens = 334800;
 export const compactionReserveTokens = contextCeilingTokens - compactionTriggerTokens;
-export const offeredGpt56Models = ["gpt-5.6-luna"];
+export const offeredOpenAiModels = hostLimitModelIds;
 class HostConfigShapeError extends Error {
     path;
     constructor(path) {
@@ -34,7 +35,7 @@ export function applyHostLimits(config) {
     const openai = optionalRecord(provider.openai, "provider.openai");
     const models = optionalRecord(openai.models, "provider.openai.models");
     const nextModels = { ...models };
-    for (const model of offeredGpt56Models) {
+    for (const model of offeredOpenAiModels) {
         const currentModel = optionalRecord(models[model], "provider.openai.models." + model);
         const limit = optionalRecord(currentModel.limit, "provider.openai.models." + model + ".limit");
         nextModels[model] = { ...currentModel, limit: { ...limit, ...modelLimit() } };

package/dist/cli/install-config.js

@@ -6,7 +6,7 @@ import { defaultGlobalConfigFile, defaultGlobalConfigJson, mergeConfigs, readLit
 import { diagnoseModelRoutes, effectiveAuthoredAgentRoutes, UnsafeModelRouteError } from "../model-route-policy.js";
 import { litOpenCodeConfigFileForOpenCodeRoot } from "../state.js";
 import { isRecord, readJsonObjectIfPresent } from "./json.js";
-import { isModelProvider, modelMenuRows, stripFastSuffix } from "./model-catalog.js";
+import { installModelEfforts, isModelProvider, modelMenuRows, stripFastSuffix } from "./model-catalog.js";
 import { classifyModelConfig, selectedLitConfig } from "./model-policy.js";
 export async function writeTextAtomically(filePath, text, overrides = {}) {
     const tempPath = path.join(path.dirname(filePath), "." + path.basename(filePath) + ".tmp-" + randomUUID());
@@ -39,7 +39,7 @@ export async function readInstallModelClassification(root) {
     return classifyModelConfig(await readValidatedConfig(root));
 }
 const installCategories = ["planning", "execution", "review", "research"];
-const installEfforts = ["low", "medium", "high", "xhigh", "max"];
+const installEfforts = installModelEfforts;
 function existingInstallRoute(value) {
     if (!isRecord(value) || typeof value.provider !== "string" || !isModelProvider(value.provider))
         return undefined;

package/dist/cli/install-report.js

@@ -1,3 +1,4 @@
+import { shippedDefaultsSummary } from "./model-catalog.js";
 const reset = "\u001b[0m";
 const green = "\u001b[38;5;82m";
 const cyan = "\u001b[1;38;5;81m";
@@ -154,7 +155,7 @@ export function renderInstallReport(report) {
         line("Permissions", permissionRoute(report)),
         line("Authority lifecycle", "schema 3; bounded history/context; trusted resume only"),
         line("Agents", "lit-loop, lit-plan, lit-implement"),
-        line("Shipped defaults", "Astra/xhigh lead + planning/review; LUNA/max execution/research helpers"),
+        line("Shipped defaults", shippedDefaultsSummary()),
         line("Override order", "project merges after global; per-agent routes override categories"),
         line("Coverage", report.litopencodeConfig.agents.length + " agents / " + report.litopencodeConfig.categories.length + " categories"),
         ...section("Host capabilities"),

package/dist/cli/managed-skill-assets.d.ts

@@ -1,5 +1,5 @@
 import type { LitOpenCodeRuntimeSkillId } from "../skills.ts";
-export type ManagedSkillId = "lit-comprehend" | "lit-handoff" | "lit-scientific-visualization" | "frontend-ui-ux" | "visual-qa" | "lit-plan" | "browser-drive" | "autoresearch" | "autoconference" | "wikify";
+export type ManagedSkillId = "lit-comprehend" | "lit-handoff" | "lit-scientific-visualization" | "readme-studio" | "frontend-ui-ux" | "visual-qa" | "lit-plan" | "browser-drive" | "autoresearch" | "autoconference" | "wikify";
 export type CanonicalManagedAsset = {
     readonly path: string;
     readonly sha256: string;

package/dist/cli/managed-skill-assets.js

@@ -10,6 +10,7 @@ export function managedSkillDefinition(id) {
         id !== "lit-handoff" &&
         id !== "lit-scientific-visualization" &&
         id !== "browser-drive" &&
+        id !== "readme-studio" &&
         id !== "frontend-ui-ux" &&
         id !== "visual-qa" &&
         id !== "lit-plan" &&
@@ -104,7 +105,8 @@ const skillDiscoveryDescriptions = Object.freeze({
     "lit-recap": "Trigger when the user asks what has happened so far — \"recap\", \"리캡\", \"where are we\", \"summarize this session\", or resuming after a break — so the answer is a read-only Korean-default digest built from the durable ledger and session evidence instead of a new action.",
     "lit-handoff": "Trigger only for explicit `/lit-handoff`, an exact bare `handoff` message, or native selection by the exact `lit-handoff` id.",
     "lit-scientific-visualization": "Trigger only for explicit `/lit-scientific-visualization`, an exact bare `lit-scientific-visualization` message, or native selection by the exact `lit-scientific-visualization` id.",
-    "frontend-ui-ux": "Trigger the moment interface work is described, not after it starts — \"design a settings page\", \"redesign the dashboard\", \"build a login screen\", or any making verb aimed at a UI, screen, layout, component, or stylesheet — so the design contract and the existing tokens shape the work before markup exists. Also on the exact `frontend-ui-ux` id. The SKILL.md reference router names which of the fifteen shipped reference documents answers which question.",
+    "readme-studio": "Select by native skill name for factual README writing and local cover composition. Inspect actual session image capability; never infer generation support or authorize publication from selection.",
+    "frontend-ui-ux": "Trigger the moment interface work is described, not after it starts — \"design a settings page\", \"redesign the dashboard\", \"build a login screen\", or any making verb aimed at a UI, screen, layout, component, or stylesheet — so compact direction and existing tokens shape the authorized build. Evolve the contract with implementation and inspect renders; review and plan intent remains read-only. Also on the exact `frontend-ui-ux` id. The SKILL.md reference router names which of the fifteen shipped reference documents answers which question.",
     "visual-qa": "Trigger when a rendered surface needs to be checked rather than described — an explicit `visual-qa` request, or verifying that an interface change actually looks and behaves right. Selection does not make a browser callable: the shipped capture playbook names the exact blocked code per unavailable channel, and a blocked check reported honestly outranks an unproven pass.",
     autoresearch: "Trigger when a measurable objective needs repeated hypothesis, experiment, evaluation, and keep/revert decisions, or when its debug, fix, learn, plan, predict, reason, scenario, security, or readiness modes fit the request.",
     autoconference: "Trigger when several genuinely independent research lanes should explore partitions and exchange peer-reviewed packets; block rather than role-play when the current root lacks real OpenCode task capability.",