superacli 1.1.2 → 1.1.4

package/cli/skills.js

@@ -1,6 +1,7 @@
 const { createPlan } = require("./planner")
 
 const PLUGINS_USAGE_SKILL_ID = "plugins.registry.usage"
+const MCP_SERVERS_USAGE_SKILL_ID = "mcp.servers.usage"
 const {
   listProviders,
   addProvider,
@@ -221,6 +222,49 @@ function buildPluginsUsageSkillMarkdown(options = {}) {
   return `---\n${renderYamlObject(frontmatter)}\n---\n\n# Instruction\n\nUse this workflow for plugin discovery and installation:\n\n1. Explore registry metadata (name/description/tags):\n\n\`\`\`bash\nsupercli plugins explore --json\nsupercli plugins explore --name commiat --json\nsupercli plugins explore --tags git,ai --json\n\`\`\`\n\n2. Install by registry name:\n\n\`\`\`bash\nsupercli plugins install commiat --json\n\`\`\`\n\n3. Install directly from a remote git repository:\n\n\`\`\`bash\nsupercli plugins install --git https://github.com/org/repo.git --manifest-path plugins/supercli/plugin.json --ref main --json\n\`\`\`\n\n4. Validate installed plugin health and guidance:\n\n\`\`\`bash\nsupercli plugins doctor commiat --json\nsupercli plugins show commiat --json\n\`\`\`\n\n5. Use the namespace command exposed by the plugin.\n\n# Examples\n\n\`\`\`bash\nsupercli skills get ${PLUGINS_USAGE_SKILL_ID} --format skill.md\nsupercli skills get ${PLUGINS_USAGE_SKILL_ID} --format skill.md --show-dag\n\`\`\``
 }
 
+function buildMcpServersUsageSkillMarkdown(options = {}) {
+  const includeDag = !!options.showDag
+  const frontmatter = {
+    skill_name: "mcp_servers_usage",
+    description: "Teaches agents how to manage and use MCP servers through SuperCLI, including browser-use style SSE bridges.",
+    command: `skills get ${MCP_SERVERS_USAGE_SKILL_ID}`,
+    arguments: [
+      {
+        name: "format",
+        type: "string",
+        required: false,
+        description: "Output format, default skill.md"
+      },
+      {
+        name: "show-dag",
+        type: "boolean",
+        required: false,
+        description: "Include internal DAG for agent reasoning"
+      }
+    ],
+    output_schema: {
+      instruction: "string",
+      examples: "array"
+    },
+    metadata: {
+      side_effects: false,
+      risk_level: "safe",
+      dag_supported: true
+    }
+  }
+
+  if (includeDag) {
+    frontmatter.dag = [
+      { id: 1, type: "list_mcp_servers" },
+      { id: 2, type: "register_or_update_mcp_server", depends_on: [1] },
+      { id: 3, type: "verify_mcp_server_configuration", depends_on: [2] },
+      { id: 4, type: "execute_mcp_backed_command", depends_on: [3] }
+    ]
+  }
+
+  return `---\n${renderYamlObject(frontmatter)}\n---\n\n# Instruction\n\nUse this workflow to manage and use MCP servers safely in non-interactive agent flows:\n\n1. List registered MCP servers and inspect current state:\n\n\`\`\`bash\nsupercli mcp list --json\n\`\`\`\n\n2. Register or update a server by URL (HTTP MCP):\n\n\`\`\`bash\nsupercli mcp add summarize-local --url http://127.0.0.1:8787 --json\n\`\`\`\n\n3. Register a stdio bridge server (browser-use style over remote SSE using mcp-remote):\n\n\`\`\`bash\n# Provide key at runtime; do not hardcode secrets in repository files\nsupercli mcp add browser-use --command npx \\\n  --args-json '["mcp-remote","https://api.browser-use.com/mcp","--header","X-Browser-Use-API-Key: \${BROWSER_USE_API_KEY}"]' \\\n  --env-json '{"BROWSER_USE_API_KEY":"\${BROWSER_USE_API_KEY}"}' \\\n  --json\n\`\`\`\n\n4. Verify the MCP server registration:\n\n\`\`\`bash\nsupercli mcp list --json\n\`\`\`\n\n5. Execute MCP-backed commands normally once configured:\n\n\`\`\`bash\n# Example command that uses adapter: mcp under the hood\nsupercli ai browser probe --json\n\`\`\`\n\n6. Remove stale MCP entries when no longer needed:\n\n\`\`\`bash\nsupercli mcp remove browser-use --json\n\`\`\`\n\n# Notes\n\n- Prefer passing secrets via runtime environment variables and CLI args, not committed files.\n- For named MCP servers, command-level adapter settings override server-level defaults on conflicts.\n- Use \`supercli skills teach --format skill.md\` for general skills discovery guidance.`
+}
+
 function listSkillsMetadata(config) {
   const commandSkills = (config.commands || []).map(cmd => ({
     name: `${cmd.namespace}.${cmd.resource}.${cmd.action}`,
@@ -230,6 +274,10 @@ function listSkillsMetadata(config) {
     name: PLUGINS_USAGE_SKILL_ID,
     description: "Discover and install plugins from the registry or remote git repos"
   })
+  commandSkills.push({
+    name: MCP_SERVERS_USAGE_SKILL_ID,
+    description: "Manage MCP servers and execute MCP-backed commands (including browser-use style bridges)"
+  })
   return commandSkills
 }
 
@@ -386,6 +434,11 @@ function handleSkillsCommand(options) {
       return true
     }
 
+    if (parsed.id === MCP_SERVERS_USAGE_SKILL_ID) {
+      console.log(buildMcpServersUsageSkillMarkdown({ showDag: !!flags["show-dag"] }))
+      return true
+    }
+
     const cmd = config.commands.find(c =>
       c.namespace === parsed.namespace && c.resource === parsed.resource && c.action === parsed.action
     )
@@ -404,10 +457,12 @@ function handleSkillsCommand(options) {
 
 module.exports = {
   PLUGINS_USAGE_SKILL_ID,
+  MCP_SERVERS_USAGE_SKILL_ID,
   normalizeSkillId,
   buildCommandSkillMarkdown,
   buildTeachSkillMarkdown,
   buildPluginsUsageSkillMarkdown,
+  buildMcpServersUsageSkillMarkdown,
   listSkillsMetadata,
   handleSkillsCommand,
   renderYamlObject // Export for testing coverage

package/cli/supercli.js

@@ -212,7 +212,7 @@ function renderTopLevelHelp(config) {
       "  Plugins: supercli plugins explore | supercli plugins install <name|path> | supercli plugins install --git <repo>",
     );
     console.log(
-      "  MCP: supercli mcp list | supercli mcp add <name> --url <url> | supercli mcp remove <name>",
+      "  MCP: supercli mcp list | supercli mcp add <name> --url <url> | supercli mcp add <name> --command <cmd> --args-json '[]' | supercli mcp remove <name>",
     );
     console.log(
       "  Skills: supercli skills list | supercli skills get <id> | supercli skills search --query <q> | supercli skills sync",

package/docs/features/adapters.md

@@ -6,8 +6,8 @@ SUPERCLI acts as a universal frontend proxy that translates semantic commands (e
 
 - **HTTP Adapter**: Directly invokes external REST APIs with configured methods, headers, and interpolated path/query parameters.
 - **OpenAPI Adapter**: Given a registered OpenAPI spec URL, maps a SUPERCLI command dynamically to an OpenAPI operation, handling auth and schema resolution on the fly.
-- **MCP Adapter (Model Context Protocol)**: Connects to local or remote MCP servers to trigger their exposed tools. Supports both standard HTTP/SSE connections and spinning up local `stdio` child process execution (like `node server.js`).
-- **Local MCP Registry**: Decouples MCP tool integration from the backend server by maintaining a local `.supercli_cache.json` registry for fast integration.
+- **MCP Adapter (Model Context Protocol)**: Connects to local or remote MCP servers to trigger their exposed tools. Supports HTTP endpoints and local `stdio` execution (`command` + `args`).
+- **Local MCP Registry**: Decouples MCP tool integration from the backend server by maintaining MCP server entries in `~/.supercli/config.json`.
 
 ## Usage
 
@@ -20,6 +20,10 @@ supercli mcp list
 # Add an offline local MCP stdio server
 supercli mcp add summarize-local --url http://127.0.0.1:8787
 
+# Add a stdio bridge for remote SSE MCP servers (browser-use style)
+supercli mcp add browser-use --command npx \
+  --args-json '["mcp-remote","https://api.browser-use.com/mcp","--header","X-Browser-Use-API-Key: your-api-key"]'
+
 # Execute an MCP tool via the bound alias
 supercli ai text summarize --text "Hello world"
 ```

package/docs/initial/mcp-local-mode.md

@@ -16,11 +16,14 @@ For `adapter: mcp`, SUPERCLI resolves the tool endpoint in this order:
 ```bash
 supercli mcp list
 supercli mcp add summarize-local --url http://127.0.0.1:8787
+supercli mcp add browser-use --command npx --args-json '["mcp-remote","https://api.browser-use.com/mcp","--header","X-Browser-Use-API-Key: your-api-key"]'
 supercli mcp remove summarize-local
 ```
 
 These commands only update local cache (`~/.supercli/config.json`).
 
+Local config also accepts a Claude-style `mcpServers` object and normalizes it to `mcp_servers` internally.
+
 ## Demo: `ai text summarize` with stdio MCP (no server)
 
 From the repository root:

package/docs/skills-catalog.md

@@ -27,6 +27,16 @@ supercli skills search --query "planning" --json
 supercli skills get opencode:plan-changes
 ```
 
+## Built-in MCP Usage Skill (Agent-Friendly)
+
+```bash
+supercli skills get mcp.servers.usage
+```
+
+This built-in skill teaches non-human agents how to register, verify, and use MCP servers
+through SuperCLI, including a browser-use style `mcp-remote` SSE bridge pattern with
+runtime API-key passing.
+
 ## Notes
 
 - IDs are always provider-qualified (`provider:id`) for future-proof disambiguation.
@@ -69,6 +79,46 @@ supercli skills get nullclaw:docs.en.commands
 project overview, operator docs, security guidance, and implementation notes on demand.
 It also exposes the local `nullclaw` binary through wrapped commands and full passthrough.
 
+## Remote Provider Example: blogwatcher
+
+```bash
+supercli plugins install blogwatcher --json
+supercli skills list --catalog --provider blogwatcher --json
+supercli skills get blogwatcher:root.skill
+supercli skills get blogwatcher:root.readme
+```
+
+`blogwatcher` indexes the upstream `SKILL.md` and `README.md` from `Hyaxia/blogwatcher`
+so agents can learn the CLI workflow, storage model, and testing expectations on demand.
+It also exposes the local `blogwatcher` binary through wrapped commands and passthrough.
+
+## Remote Provider Example: xurl
+
+```bash
+supercli plugins install xurl --json
+supercli skills list --catalog --provider xurl --json
+supercli skills get xurl:root.skill
+supercli skills get xurl:root.readme
+```
+
+`xurl` indexes the upstream `SKILL.md` and `README.md` from `xdevplatform/xurl`
+so agents can learn safe X API workflows, auth constraints, and shortcut command patterns
+without reading local token storage. It also exposes curated read-only wrappers for the local
+`xurl` binary.
+
+## Remote Provider Example: clix
+
+```bash
+supercli plugins install clix --json
+supercli skills list --catalog --provider clix --json
+supercli skills get clix:root.skill
+supercli skills get clix:root.readme
+```
+
+`clix` indexes the upstream `SKILL.md` and `README.md` from `spideystreet/clix`
+so agents can learn safe cookie-auth X workflows, JSON usage, and agent-mode guidance.
+It also exposes curated read-only wrappers for the local `clix` binary.
+
 ## Local Repo Skill Example: cline-non-interactive
 
 ```bash

package/docs/supported-harnesses.md

@@ -209,6 +209,26 @@ These harnesses are in active development and will be available as plugins. Stat
 
 ---
 
+### Databases
+
+| Harness | CLI | Status | Purpose |
+|---------|-----|--------|---------|
+| **mysql** | mysql | Available via plugin | MySQL client version checks, one-off queries, passthrough |
+| **mongosh** | mongosh | Available via plugin | MongoDB shell ping, eval, and passthrough |
+
+---
+
+### Communication
+
+| Harness | CLI | Status | Purpose |
+|---------|-----|--------|---------|
+| **himalaya** | himalaya | Available via plugin | Read-only email account, folder, envelope, and preview workflows |
+| **wacli** | wacli | Available via plugin | Read-only WhatsApp diagnostics, chats, messages, contacts, and groups |
+| **xurl** | xurl | Available via plugin | Read-only X account, timeline, search, and social graph workflows |
+| **clix** | clix | Available via plugin | Read-only X timeline, search, user, tweet, and bookmarks workflows |
+
+---
+
 ### Version Control & Release
 
 | Harness | CLI | Status | Purpose |

package/package.json

@@ -1,7 +1,8 @@
 {
   "name": "superacli",
-  "version": "1.1.2",
+  "version": "1.1.4",
   "description": "Config-driven, AI-friendly dynamic CLI",
+  "license": "MIT",
   "main": "server/app.js",
   "bin": {
     "dcli": "./cli/supercli.js",

package/plugins/blogwatcher/README.md

@@ -0,0 +1,52 @@
+# BlogWatcher Plugin
+
+This plugin is a hybrid harness: it indexes BlogWatcher's upstream `SKILL.md` and `README.md` into the local skills catalog and exposes the local `blogwatcher` binary through wrapped commands plus full passthrough.
+
+## What It Adds
+
+- agent-oriented usage guidance from the upstream `SKILL.md`
+- project and end-user documentation from the upstream `README.md`
+- non-interactive wrappers for the local BlogWatcher CLI
+
+## Install
+
+```bash
+supercli plugins install blogwatcher --json
+```
+
+## Explore Indexed Skills
+
+```bash
+supercli skills list --catalog --provider blogwatcher --json
+supercli skills get blogwatcher:root.skill
+supercli skills get blogwatcher:root.readme
+```
+
+## Available CLI Commands
+
+```bash
+# Wrapped version command
+supercli blogwatcher cli version --json
+
+# Wrapped list/add/remove commands
+supercli blogwatcher blogs list --json
+supercli blogwatcher blogs add --name "Example" --url "https://example.com/blog" --feed-url "https://example.com/feed.xml" --json
+supercli blogwatcher blogs remove --name "Example" --json
+
+# Wrapped scan and article commands
+supercli blogwatcher scan run --workers 4 --json
+supercli blogwatcher articles list --all --json
+supercli blogwatcher articles read-all --json
+
+# Full passthrough
+supercli blogwatcher articles --all
+supercli blogwatcher scan "Example Blog"
+```
+
+## Notes
+
+- This plugin indexes remote markdown from `https://github.com/Hyaxia/blogwatcher`.
+- Wrapped commands return raw text in the dcli envelope because upstream does not expose JSON output.
+- The wrapped `blogs remove` and `articles read-all` commands force `--yes` to avoid interactive prompts.
+- BlogWatcher stores its SQLite data under `~/.blogwatcher`, so use an isolated `HOME` if you want disposable smoke testing.
+- Removing the plugin also removes its registered skills provider from the local catalog.

package/plugins/blogwatcher/plugin.json

@@ -0,0 +1,195 @@
+{
+  "name": "blogwatcher",
+  "version": "0.1.0",
+  "description": "Hybrid BlogWatcher plugin with remote agent docs plus local CLI wrappers",
+  "source": "https://github.com/Hyaxia/blogwatcher",
+  "checks": [
+    { "type": "binary", "name": "curl" },
+    { "type": "binary", "name": "blogwatcher" }
+  ],
+  "post_install": {
+    "script": "scripts/post-install.js",
+    "runtime": "node",
+    "timeout_ms": 15000
+  },
+  "post_uninstall": {
+    "script": "scripts/post-uninstall.js",
+    "runtime": "node",
+    "timeout_ms": 15000
+  },
+  "commands": [
+    {
+      "namespace": "blogwatcher",
+      "resource": "cli",
+      "action": "version",
+      "description": "Show BlogWatcher CLI version",
+      "adapter": "process",
+      "adapterConfig": {
+        "command": "blogwatcher",
+        "baseArgs": ["--version"],
+        "parseJson": false,
+        "missingDependencyHelp": "Please install blogwatcher to use this command."
+      },
+      "args": []
+    },
+    {
+      "namespace": "blogwatcher",
+      "resource": "blogs",
+      "action": "list",
+      "description": "List tracked blogs",
+      "adapter": "process",
+      "adapterConfig": {
+        "command": "blogwatcher",
+        "baseArgs": ["blogs"],
+        "parseJson": false,
+        "timeout_ms": 15000,
+        "missingDependencyHelp": "Please install blogwatcher to use this command."
+      },
+      "args": []
+    },
+    {
+      "namespace": "blogwatcher",
+      "resource": "blogs",
+      "action": "add",
+      "description": "Add a blog to BlogWatcher",
+      "adapter": "process",
+      "adapterConfig": {
+        "command": "blogwatcher",
+        "baseArgs": ["add"],
+        "positionalArgs": ["name", "url"],
+        "parseJson": false,
+        "timeout_ms": 15000,
+        "missingDependencyHelp": "Please install blogwatcher to use this command."
+      },
+      "args": [
+        { "name": "name", "type": "string", "required": true },
+        { "name": "url", "type": "string", "required": true },
+        { "name": "feed-url", "type": "string", "required": false },
+        { "name": "scrape-selector", "type": "string", "required": false }
+      ]
+    },
+    {
+      "namespace": "blogwatcher",
+      "resource": "blogs",
+      "action": "remove",
+      "description": "Remove a tracked blog without prompting",
+      "adapter": "process",
+      "adapterConfig": {
+        "command": "blogwatcher",
+        "baseArgs": ["remove", "--yes"],
+        "positionalArgs": ["name"],
+        "parseJson": false,
+        "timeout_ms": 15000,
+        "missingDependencyHelp": "Please install blogwatcher to use this command."
+      },
+      "args": [
+        { "name": "name", "type": "string", "required": true }
+      ]
+    },
+    {
+      "namespace": "blogwatcher",
+      "resource": "scan",
+      "action": "run",
+      "description": "Scan blogs for new articles in non-interactive mode",
+      "adapter": "process",
+      "adapterConfig": {
+        "command": "blogwatcher",
+        "baseArgs": ["scan", "--silent"],
+        "positionalArgs": ["blog"],
+        "parseJson": false,
+        "timeout_ms": 15000,
+        "missingDependencyHelp": "Please install blogwatcher to use this command."
+      },
+      "args": [
+        { "name": "blog", "type": "string", "required": false },
+        { "name": "workers", "type": "integer", "required": false }
+      ]
+    },
+    {
+      "namespace": "blogwatcher",
+      "resource": "articles",
+      "action": "list",
+      "description": "List tracked articles",
+      "adapter": "process",
+      "adapterConfig": {
+        "command": "blogwatcher",
+        "baseArgs": ["articles"],
+        "parseJson": false,
+        "timeout_ms": 15000,
+        "missingDependencyHelp": "Please install blogwatcher to use this command."
+      },
+      "args": [
+        { "name": "all", "type": "boolean", "required": false },
+        { "name": "blog", "type": "string", "required": false }
+      ]
+    },
+    {
+      "namespace": "blogwatcher",
+      "resource": "articles",
+      "action": "read",
+      "description": "Mark an article as read",
+      "adapter": "process",
+      "adapterConfig": {
+        "command": "blogwatcher",
+        "baseArgs": ["read"],
+        "positionalArgs": ["article_id"],
+        "parseJson": false,
+        "timeout_ms": 15000,
+        "missingDependencyHelp": "Please install blogwatcher to use this command."
+      },
+      "args": [
+        { "name": "article_id", "type": "integer", "required": true }
+      ]
+    },
+    {
+      "namespace": "blogwatcher",
+      "resource": "articles",
+      "action": "unread",
+      "description": "Mark an article as unread",
+      "adapter": "process",
+      "adapterConfig": {
+        "command": "blogwatcher",
+        "baseArgs": ["unread"],
+        "positionalArgs": ["article_id"],
+        "parseJson": false,
+        "timeout_ms": 15000,
+        "missingDependencyHelp": "Please install blogwatcher to use this command."
+      },
+      "args": [
+        { "name": "article_id", "type": "integer", "required": true }
+      ]
+    },
+    {
+      "namespace": "blogwatcher",
+      "resource": "articles",
+      "action": "read-all",
+      "description": "Mark unread articles as read without prompting",
+      "adapter": "process",
+      "adapterConfig": {
+        "command": "blogwatcher",
+        "baseArgs": ["read-all", "--yes"],
+        "parseJson": false,
+        "timeout_ms": 15000,
+        "missingDependencyHelp": "Please install blogwatcher to use this command."
+      },
+      "args": [
+        { "name": "blog", "type": "string", "required": false }
+      ]
+    },
+    {
+      "namespace": "blogwatcher",
+      "resource": "_",
+      "action": "_",
+      "description": "Passthrough command to execute any BlogWatcher CLI command",
+      "adapter": "process",
+      "adapterConfig": {
+        "command": "blogwatcher",
+        "passthrough": true,
+        "parseJson": false,
+        "timeout_ms": 15000,
+        "missingDependencyHelp": "Please install blogwatcher to use this command."
+      },
+      "args": []
+    }
+  ]
+}

package/plugins/blogwatcher/scripts/post-install.js

@@ -0,0 +1,66 @@
+const { addProvider, syncCatalog } = require("../../../cli/skills-catalog")
+
+const OWNER = "Hyaxia"
+const REPO = "blogwatcher"
+const REF = "main"
+const SOURCE_REPO = `https://github.com/${OWNER}/${REPO}`
+const RAW_BASE_URL = `https://raw.githubusercontent.com/${OWNER}/${REPO}/${REF}`
+
+const CATALOG_FILES = [
+  {
+    id: "root.skill",
+    name: "BlogWatcher Agent Skill",
+    path: "SKILL.md",
+    description: "Agent-focused guidance for working with the BlogWatcher CLI, code layout, and testing flow.",
+    tags: ["agents", "blogs", "workflow"]
+  },
+  {
+    id: "root.readme",
+    name: "BlogWatcher Overview",
+    path: "README.md",
+    description: "Project overview, installation steps, storage model, and end-user command examples.",
+    tags: ["overview", "blogs", "rss"]
+  }
+]
+
+function buildRemoteEntries() {
+  return CATALOG_FILES.map(file => ({
+    ...file,
+    source_url: `${RAW_BASE_URL}/${file.path}`
+  }))
+}
+
+function run() {
+  const entries = buildRemoteEntries()
+  addProvider({
+    name: "blogwatcher",
+    type: "remote_static",
+    enabled: true,
+    source_repo: SOURCE_REPO,
+    ref: REF,
+    entries
+  })
+
+  const index = syncCatalog()
+  return {
+    provider: "blogwatcher",
+    entries: entries.length,
+    synced_skills: Array.isArray(index.skills) ? index.skills.length : 0
+  }
+}
+
+if (require.main === module) {
+  try {
+    const result = run()
+    process.stdout.write(JSON.stringify(result))
+  } catch (err) {
+    process.stderr.write(err.message)
+    process.exit(1)
+  }
+}
+
+module.exports = {
+  CATALOG_FILES,
+  buildRemoteEntries,
+  run
+}

package/plugins/blogwatcher/scripts/post-uninstall.js

@@ -0,0 +1,25 @@
+const { removeProvider, syncCatalog } = require("../../../cli/skills-catalog")
+
+function run() {
+  const removed = removeProvider("blogwatcher")
+  const index = syncCatalog()
+  return {
+    provider: "blogwatcher",
+    removed,
+    synced_skills: Array.isArray(index.skills) ? index.skills.length : 0
+  }
+}
+
+if (require.main === module) {
+  try {
+    const result = run()
+    process.stdout.write(JSON.stringify(result))
+  } catch (err) {
+    process.stderr.write(err.message)
+    process.exit(1)
+  }
+}
+
+module.exports = {
+  run
+}

package/plugins/clix/README.md

@@ -0,0 +1,44 @@
+# clix Plugin Harness
+
+This plugin is a hybrid harness for `clix`: it indexes the upstream `SKILL.md` and `README.md` into the local skills catalog and exposes a curated set of safe read-only wrappers.
+
+## Why This Scope
+
+`clix` can post, delete, like, retweet, and mutate bookmarks using cookie-based auth. The wrapped commands in this plugin intentionally focus on low-risk read-only workflows.
+
+- upstream agent docs are indexed into the skills catalog
+- read-only wrappers are exposed for auth status, timelines, search, tweets, users, and bookmarks
+- no wildcard passthrough in v1
+- no auth mutation, posting, delete, like, retweet, or bookmark mutation wrappers
+
+## Install
+
+```bash
+supercli plugins install clix --json
+```
+
+## Explore Indexed Skills
+
+```bash
+supercli skills list --catalog --provider clix --json
+supercli skills get clix:root.skill
+supercli skills get clix:root.readme
+```
+
+## Available CLI Commands
+
+```bash
+supercli clix auth status --json
+supercli clix timeline list --type following --count 20 --json
+supercli clix posts search --query "from:openai" --type latest --count 20 --json
+supercli clix posts show --id 1234567890 --json
+supercli clix users show --handle openai --json
+supercli clix bookmarks list --json
+```
+
+## Notes
+
+- Wrapped commands always request upstream `--json` output and return parsed data in the dcli envelope.
+- `clix` uses cookie-based auth and may access local browser/session data. Treat that auth state as sensitive.
+- Use upstream `clix` directly for login, account switching, posting, deleting, likes, retweets, and bookmark mutations.
+- Removing the plugin also removes its registered skills provider from the local catalog.