Openclaw local plugin integration (#1488)

## Description

This PR integrates recent fixes to improve skill generation and handle
viewer authentication gracefully.

- **Skill Generator**: Aligned the skill generation prompt template with
skill-creator best practices. Added explicit instructions to assume the
agent is smart, improved trigger descriptions, enforced stricter
metadata/naming rules, and prevented the generation of extraneous
boilerplate files.
- **Viewer**: Added a check in the `loadConfig` function to show a
"session expired" toast notification when receiving a 401 Unauthorized
response, preventing silent authentication failures.

Related Issue (Required):  Fixes #1434, Fixes #1424

## Type of change

Please delete options that are not relevant.

- [x] Bug fix (non-breaking change which fixes an issue)
- [ ] New feature (non-breaking change which adds functionality)
- [ ] Breaking change (fix or feature that would cause existing
functionality to not work as expected)
- [ ] Refactor (does not change functionality, e.g. code style
improvements, linting)
- [ ] Documentation update

## How Has This Been Tested?

Please describe the tests that you ran to verify your changes. Provide
instructions so we can reproduce. Please also list any relevant details
for your test configuration

- [ ] Unit Test
- [x] Test Script Or Test Steps (please provide)
- Generated a new skill to verify that the output correctly follows the
updated prompt constraints (e.g., kebab-case naming, no extra files,
proper trigger keywords).
- Manually simulated a 401 response during config load in the viewer to
ensure the session expiration toast appears properly.
- [ ] Pipeline Automated API Test (please provide)

## Checklist

- [x] I have performed a self-review of my own code | 我已自行检查了自己的代码
- [x] I have commented my code in hard-to-understand areas |
我已在难以理解的地方对代码进行了注释
- [x] I have added tests that prove my fix is effective or that my
feature works | 我已添加测试以证明我的修复有效或功能正常
- [ ] I have created related documentation issue/PR in
[MemOS-Docs](https://github.com/MemTensor/MemOS-Docs) (if applicable) |
我已在 [MemOS-Docs](https://github.com/MemTensor/MemOS-Docs) 中创建了相关的文档
issue/PR（如果适用）
- [x] I have linked the issue to this PR (if applicable) | 我已将 issue
链接到此 PR（如果适用）
- [ ] I have mentioned the person who will review this PR | 我已提及将审查此 PR
的人

## Reviewer Checklist
- [x] closes #1434, closes #1424
- [ ] Made sure Checks passed
- [ ] Tests have been provided

Author: Hun-ger

apps/memos-local-openclaw/src/skill/generator.ts

@@ -24,17 +24,24 @@ This Skill is special: it comes from real execution experience — every step wa
 
 ## Core principles (follow strictly but do NOT include these in output)
 
+### Assume the agent is smart
+The agent consuming this skill is an LLM — it already knows how to use git, write code, read docs, and run commands.
+- Only write what the agent CANNOT derive from the codebase or general knowledge: project-specific gotchas, non-obvious config, verified flags/versions, and real pitfalls encountered.
+- Do NOT explain basic concepts, standard tool usage, or things any competent developer already knows.
+
 ### Progressive disclosure
 - The frontmatter description (~100 words) is ALWAYS in the agent's context — it must be self-sufficient for deciding whether to use this skill.
 - The SKILL.md body loads when triggered — keep it under 400 lines, focused, no fluff.
 - If the task involved large configs/scripts, mention them but DON'T inline everything — just reference that scripts/ or references/ may contain them.
+- If the skill covers multiple variants (e.g. different OS targets, different cloud providers), put variant-specific details in references/ files. The main SKILL.md should cover the common workflow only.
 
 ### Description as trigger mechanism
-The description field decides whether the agent activates this skill. Write it "proactively":
+The description field is the SOLE input the agent uses to decide whether to activate this skill. If the description doesn't match, the body is never read — so put maximum effort here.
 - Don't just say what it does — list the situations, keywords, and phrasings that should trigger it.
 - Claude/agents tend to under-trigger skills. Counter this by being explicit about when to use it.
+- Include indirect phrasings: users often describe the goal ("make it run anywhere") rather than the tool ("Docker").
 - Bad: "How to deploy Node.js to Docker"
-- Good: "How to containerize and deploy a Node.js application using Docker. Use when the user mentions Docker deployment, Dockerfile writing, container builds, multi-stage builds, port mapping, .dockerignore, image optimization, CI/CD container pipelines, or any task involving packaging a Node/JS backend into a container — even if they don't say 'Docker' explicitly but describe wanting to 'package the app for production' or 'run it anywhere'."
+- Good: "How to containerize and deploy a Node.js application using Docker. TRIGGER when: user mentions Docker deployment, Dockerfile writing, container builds, multi-stage builds, port mapping, .dockerignore, image optimization, CI/CD container pipelines, or any task involving packaging a Node/JS backend into a container — even if they don't say 'Docker' explicitly but describe wanting to 'package the app for production' or 'run it anywhere'."
 
 ### Writing style
 - Use imperative form
@@ -43,6 +50,9 @@ The description field decides whether the agent activates this skill. Write it "
 - Generalize from the specific task so the skill works for similar future scenarios, don't over-fit to this exact project
 - Keep real commands/code/config from the task record — these are verified to work
 
+### No extra files
+- Output ONLY the SKILL.md content. Do NOT create or reference README.md, CHANGELOG.md, INSTALLATION_GUIDE.md, CONTRIBUTING.md, or similar boilerplate files. All information belongs in SKILL.md, scripts/, or references/.
+
 ### Language matching (CRITICAL)
 You MUST write the ENTIRE skill in the SAME language as the user's messages in the task record.
 - If the user wrote in Chinese → the skill title, description, all prose sections MUST be in Chinese
@@ -57,12 +67,16 @@ DO NOT default to English. Look at the task record below and match its language.
 
 Output ONLY the complete SKILL.md content. No extra text before or after.
 
+The frontmatter contains ONLY name and description. Nothing else.
+(OpenClaw metadata such as emoji is optional — if desired, add it as a \`<!-- metadata: {{"openclaw": {{"emoji": "..."}}}}} -->\` HTML comment at the very end of the file body, NOT in the frontmatter.)
+
 ---
 name: "{NAME}"
-description: "{A natural, proactive description. 60-120 words. Cover what it does + multiple phrasings/scenarios that should trigger it. Be pushy about triggering — list keywords, alternative descriptions, edge-case phrasings.}"
-metadata: {{ "openclaw": {{ "emoji": "{emoji}" }} }}
+description: "{A natural, proactive description. 60-120 words. Cover what it does + multiple phrasings/scenarios that should trigger it. Include a TRIGGER line listing keywords and indirect phrasings. Be pushy about triggering.}"
 ---
 
+Name rules: 2-4 English words, kebab-case, lowercase (e.g. "docker-node-deploy", "aws-s3-backup"). This is a machine identifier.
+
 # {Title — clear, action-oriented}
 
 {One sentence: what this skill helps you do and why it's valuable}

apps/memos-local-openclaw/src/viewer/html.ts

@@ -7205,6 +7205,7 @@ function onProviderChange(){}
 async function loadConfig(){
   try{
     const r=await fetch('/api/config');
+    if(r.status===401){toast(t('settings.session.expired'),'error');return;}
     if(!r.ok) return;
     const cfg=await r.json();
     const emb=cfg.embedding||{};