incur 0.2.2 → 0.3.0

package/src/Help.test.ts

@@ -28,7 +28,7 @@ describe('formatCommand', () => {
         --filter-output <keys>              Filter output by key paths (e.g. foo,bar.baz,a[0,3])
         --format <toon|json|yaml|md|jsonl>  Output format
         --help                              Show help
-        --llms                              Print LLM-readable manifest
+        --llms, --llms-full                 Print LLM-readable manifest
         --schema                            Show JSON Schema for a command
         --token-count                       Print token count of output (instead of output)
         --token-limit <n>                   Limit output to n tokens
@@ -50,7 +50,7 @@ describe('formatCommand', () => {
         --filter-output <keys>              Filter output by key paths (e.g. foo,bar.baz,a[0,3])
         --format <toon|json|yaml|md|jsonl>  Output format
         --help                              Show help
-        --llms                              Print LLM-readable manifest
+        --llms, --llms-full                 Print LLM-readable manifest
         --schema                            Show JSON Schema for a command
         --token-count                       Print token count of output (instead of output)
         --token-limit <n>                   Limit output to n tokens
@@ -79,7 +79,7 @@ describe('formatCommand', () => {
         --filter-output <keys>              Filter output by key paths (e.g. foo,bar.baz,a[0,3])
         --format <toon|json|yaml|md|jsonl>  Output format
         --help                              Show help
-        --llms                              Print LLM-readable manifest
+        --llms, --llms-full                 Print LLM-readable manifest
         --schema                            Show JSON Schema for a command
         --token-count                       Print token count of output (instead of output)
         --token-limit <n>                   Limit output to n tokens
@@ -125,7 +125,7 @@ describe('formatRoot', () => {
         --filter-output <keys>              Filter output by key paths (e.g. foo,bar.baz,a[0,3])
         --format <toon|json|yaml|md|jsonl>  Output format
         --help                              Show help
-        --llms                              Print LLM-readable manifest
+        --llms, --llms-full                 Print LLM-readable manifest
         --schema                            Show JSON Schema for a command
         --token-count                       Print token count of output (instead of output)
         --token-limit <n>                   Limit output to n tokens
@@ -150,7 +150,7 @@ describe('formatRoot', () => {
         --filter-output <keys>              Filter output by key paths (e.g. foo,bar.baz,a[0,3])
         --format <toon|json|yaml|md|jsonl>  Output format
         --help                              Show help
-        --llms                              Print LLM-readable manifest
+        --llms, --llms-full                 Print LLM-readable manifest
         --schema                            Show JSON Schema for a command
         --token-count                       Print token count of output (instead of output)
         --token-limit <n>                   Limit output to n tokens
@@ -179,7 +179,7 @@ describe('formatRoot', () => {
         --filter-output <keys>              Filter output by key paths (e.g. foo,bar.baz,a[0,3])
         --format <toon|json|yaml|md|jsonl>  Output format
         --help                              Show help
-        --llms                              Print LLM-readable manifest
+        --llms, --llms-full                 Print LLM-readable manifest
         --schema                            Show JSON Schema for a command
         --token-count                       Print token count of output (instead of output)
         --token-limit <n>                   Limit output to n tokens
@@ -208,7 +208,7 @@ describe('formatRoot', () => {
         --filter-output <keys>              Filter output by key paths (e.g. foo,bar.baz,a[0,3])
         --format <toon|json|yaml|md|jsonl>  Output format
         --help                              Show help
-        --llms                              Print LLM-readable manifest
+        --llms, --llms-full                 Print LLM-readable manifest
         --schema                            Show JSON Schema for a command
         --token-count                       Print token count of output (instead of output)
         --token-limit <n>                   Limit output to n tokens

package/src/Help.ts

@@ -329,7 +329,7 @@ function globalOptionsLines(root = false): string[] {
     { flag: '--filter-output <keys>', desc: 'Filter output by key paths (e.g. foo,bar.baz,a[0,3])' },
     { flag: '--format <toon|json|yaml|md|jsonl>', desc: 'Output format' },
     { flag: '--help', desc: 'Show help' },
-    { flag: '--llms', desc: 'Print LLM-readable manifest' },
+    { flag: '--llms, --llms-full', desc: 'Print LLM-readable manifest' },
     ...(root ? [{ flag: '--mcp', desc: 'Start as MCP stdio server' }] : []),
     { flag: '--schema', desc: 'Show JSON Schema for a command' },
     { flag: '--token-count', desc: 'Print token count of output (instead of output)' },

package/src/Skill.test.ts

@@ -145,6 +145,41 @@ test('concatenates multiple commands', () => {
   `)
 })
 
+describe('index', () => {
+  test('generates compact command index', () => {
+    const result = Skill.index('test', [
+      { name: 'ping', description: 'Health check' },
+      { name: 'greet', description: 'Greet someone', args: z.object({ name: z.string() }) },
+    ])
+    expect(result).toMatchInlineSnapshot(`
+      "# test
+
+      | Command | Description |
+      |---------|-------------|
+      | \`test ping\` | Health check |
+      | \`test greet <name>\` | Greet someone |
+
+      Run \`test --llms-full\` for full manifest. Run \`test <command> --schema\` for argument details."
+    `)
+  })
+
+  test('uses brackets for optional args', () => {
+    const result = Skill.index('test', [
+      {
+        name: 'install',
+        description: 'Install a package',
+        args: z.object({ package: z.string().optional() }),
+      },
+    ])
+    expect(result).toContain('`test install [package]`')
+  })
+
+  test('handles commands without descriptions', () => {
+    const result = Skill.index('test', [{ name: 'ping' }])
+    expect(result).toContain('| `test ping` |  |')
+  })
+})
+
 describe('hash', () => {
   test('returns consistent hash for same commands', () => {
     const commands: Skill.CommandInfo[] = [
@@ -218,6 +253,7 @@ describe('split', () => {
       "---
       name: gh-auth
       description: Authenticate with GitHub. Log in, Check status. Run \`gh auth --help\` for usage details.
+      requires_bin: gh
       command: gh auth
       ---
 
@@ -235,6 +271,7 @@ describe('split', () => {
       "---
       name: gh-pr
       description: Manage pull requests. List PRs, Create PR. Run \`gh pr --help\` for usage details.
+      requires_bin: gh
       command: gh pr
       ---
 
@@ -294,9 +331,16 @@ describe('split', () => {
     expect(files[0]!.content).toContain('Run `gh auth login --help` for usage details.')
   })
 
-  test('omits --help hint when no descriptions exist', () => {
+  test('emits fallback description when no explicit descriptions exist', () => {
     const files = Skill.split('test', [{ name: 'ping' }], 1)
-    expect(files[0]!.content).not.toContain('--help')
+    expect(files[0]!.content).toContain(
+      'description: Run `test ping --help` for usage details.',
+    )
+  })
+
+  test('includes requires_bin in frontmatter', () => {
+    const files = Skill.split('gh', [{ name: 'auth login', description: 'Log in' }], 1)
+    expect(files[0]!.content).toContain('requires_bin: gh')
   })
 
   test('no per-command frontmatter in split files', () => {

package/src/Skill.ts

@@ -23,6 +23,33 @@ export type File = {
   content: string
 }
 
+/** Generates a compact Markdown command index for `--llms`. */
+export function index(name: string, commands: CommandInfo[], description?: string | undefined): string {
+  const lines: string[] = [`# ${name}`]
+  if (description) lines.push('', description)
+  lines.push('')
+  lines.push('| Command | Description |')
+  lines.push('|---------|-------------|')
+  for (const cmd of commands) {
+    const signature = buildSignature(name, cmd)
+    const desc = cmd.description ?? ''
+    lines.push(`| \`${signature}\` | ${desc} |`)
+  }
+  lines.push('', `Run \`${name} --llms-full\` for full manifest. Run \`${name} <command> --schema\` for argument details.`)
+  return lines.join('\n')
+}
+
+/** @internal Builds a command signature with arg placeholders. */
+function buildSignature(cli: string, cmd: CommandInfo): string {
+  const base = `${cli} ${cmd.name}`
+  if (!cmd.args) return base
+  const shape = cmd.args.shape as Record<string, z.ZodType>
+  const json = Schema.toJsonSchema(cmd.args)
+  const required = new Set((json.required as string[] | undefined) ?? [])
+  const argNames = Object.keys(shape).map((k) => (required.has(k) ? `<${k}>` : `[${k}]`))
+  return `${base} ${argNames.join(' ')}`
+}
+
 /** Generates a Markdown skill file from a CLI name and collected command data. */
 export function generate(
   name: string,
@@ -88,12 +115,15 @@ function renderGroup(
   const descParts: string[] = []
   if (groupDesc) descParts.push(groupDesc.replace(/\.$/, ''))
   if (childDescs.length > 0) descParts.push(childDescs.join(', '))
-  const description = descParts.join('. ') || undefined
+  const description =
+    descParts.length > 0
+      ? `${descParts.join('. ')}. Run \`${title} --help\` for usage details.`
+      : `Run \`${title} --help\` for usage details.`
 
   const slug = title.replace(/\s+/g, '-')
   const fm = ['---', `name: ${slug}`]
-  if (description)
-    fm.push(`description: ${description}. Run \`${title} --help\` for usage details.`)
+  fm.push(`description: ${description}`)
+  fm.push(`requires_bin: ${cli}`)
   fm.push(`command: ${title}`, '---')
 
   const body = cmds.map((cmd) => renderCommandBody(cli, cmd)).join('\n\n---\n\n')

package/src/SyncSkills.ts

@@ -13,8 +13,8 @@ export async function sync(
   commands: Map<string, any>,
   options: sync.Options = {},
 ): Promise<sync.Result> {
-  const cwd = options.cwd ?? resolvePackageRoot()
   const { depth = 1, description, global = true } = options
+  const cwd = options.cwd ?? (global ? resolvePackageRoot() : process.cwd())
 
   const groups = new Map<string, string>()
   if (description) groups.set(name, description)

package/src/e2e.test.ts

@@ -939,7 +939,7 @@ describe('help', () => {
         --filter-output <keys>              Filter output by key paths (e.g. foo,bar.baz,a[0,3])
         --format <toon|json|yaml|md|jsonl>  Output format
         --help                              Show help
-        --llms                              Print LLM-readable manifest
+        --llms, --llms-full                 Print LLM-readable manifest
         --mcp                               Start as MCP stdio server
         --schema                            Show JSON Schema for a command
         --token-count                       Print token count of output (instead of output)
@@ -973,7 +973,7 @@ describe('help', () => {
         --filter-output <keys>              Filter output by key paths (e.g. foo,bar.baz,a[0,3])
         --format <toon|json|yaml|md|jsonl>  Output format
         --help                              Show help
-        --llms                              Print LLM-readable manifest
+        --llms, --llms-full                 Print LLM-readable manifest
         --schema                            Show JSON Schema for a command
         --token-count                       Print token count of output (instead of output)
         --token-limit <n>                   Limit output to n tokens
@@ -1000,7 +1000,7 @@ describe('help', () => {
         --filter-output <keys>              Filter output by key paths (e.g. foo,bar.baz,a[0,3])
         --format <toon|json|yaml|md|jsonl>  Output format
         --help                              Show help
-        --llms                              Print LLM-readable manifest
+        --llms, --llms-full                 Print LLM-readable manifest
         --schema                            Show JSON Schema for a command
         --token-count                       Print token count of output (instead of output)
         --token-limit <n>                   Limit output to n tokens
@@ -1026,7 +1026,7 @@ describe('help', () => {
         --filter-output <keys>              Filter output by key paths (e.g. foo,bar.baz,a[0,3])
         --format <toon|json|yaml|md|jsonl>  Output format
         --help                              Show help
-        --llms                              Print LLM-readable manifest
+        --llms, --llms-full                 Print LLM-readable manifest
         --schema                            Show JSON Schema for a command
         --token-count                       Print token count of output (instead of output)
         --token-limit <n>                   Limit output to n tokens
@@ -1058,7 +1058,7 @@ describe('help', () => {
         --filter-output <keys>              Filter output by key paths (e.g. foo,bar.baz,a[0,3])
         --format <toon|json|yaml|md|jsonl>  Output format
         --help                              Show help
-        --llms                              Print LLM-readable manifest
+        --llms, --llms-full                 Print LLM-readable manifest
         --schema                            Show JSON Schema for a command
         --token-count                       Print token count of output (instead of output)
         --token-limit <n>                   Limit output to n tokens
@@ -1090,9 +1090,9 @@ describe('help', () => {
   })
 })
 
-describe('--llms', () => {
+describe('--llms-full', () => {
   test('json manifest lists all leaf commands sorted', async () => {
-    const { output } = await serve(createApp(), ['--llms', '--format', 'json'])
+    const { output } = await serve(createApp(), ['--llms-full', '--format', 'json'])
     const manifest = json(output)
     expect(manifest.version).toBe('incur.v1')
     const names = manifest.commands.map((c: any) => c.name)
@@ -1127,7 +1127,7 @@ describe('--llms', () => {
   })
 
   test('manifest includes schema.args and schema.options separately', async () => {
-    const { output } = await serve(createApp(), ['--llms', '--format', 'json'])
+    const { output } = await serve(createApp(), ['--llms-full', '--format', 'json'])
     const projectList = json(output).commands.find((c: any) => c.name === 'project list')
     expect(projectList.schema.options.properties).toMatchInlineSnapshot(`
       {
@@ -1157,7 +1157,7 @@ describe('--llms', () => {
   })
 
   test('manifest includes schema.output', async () => {
-    const { output } = await serve(createApp(), ['--llms', '--format', 'json'])
+    const { output } = await serve(createApp(), ['--llms-full', '--format', 'json'])
     const projectGet = json(output).commands.find((c: any) => c.name === 'project get')
     expect(projectGet.schema.output).toMatchInlineSnapshot(`
       {
@@ -1204,13 +1204,13 @@ describe('--llms', () => {
   })
 
   test('manifest omits schema when no schemas defined', async () => {
-    const { output } = await serve(createApp(), ['--llms', '--format', 'json'])
+    const { output } = await serve(createApp(), ['--llms-full', '--format', 'json'])
     const ping = json(output).commands.find((c: any) => c.name === 'ping')
     expect(ping.schema).toBeUndefined()
   })
 
-  test('scoped --llms to group', async () => {
-    const { output } = await serve(createApp(), ['auth', '--llms', '--format', 'json'])
+  test('scoped --llms-full to group', async () => {
+    const { output } = await serve(createApp(), ['auth', '--llms-full', '--format', 'json'])
     const names = json(output).commands.map((c: any) => c.name)
     expect(names).toMatchInlineSnapshot(`
       [
@@ -1221,8 +1221,8 @@ describe('--llms', () => {
     `)
   })
 
-  test('scoped --llms to nested group', async () => {
-    const { output } = await serve(createApp(), ['project', 'deploy', '--llms', '--format', 'json'])
+  test('scoped --llms-full to nested group', async () => {
+    const { output } = await serve(createApp(), ['project', 'deploy', '--llms-full', '--format', 'json'])
     const names = json(output).commands.map((c: any) => c.name)
     expect(names).toMatchInlineSnapshot(`
       [
@@ -1233,27 +1233,27 @@ describe('--llms', () => {
     `)
   })
 
-  test('default --llms outputs markdown', async () => {
-    const { output } = await serve(createApp(), ['--llms'])
+  test('default --llms-full outputs markdown', async () => {
+    const { output } = await serve(createApp(), ['--llms-full'])
     expect(output).toContain('# app')
     expect(output).toContain('auth login')
     expect(output).toContain('project list')
   })
 
-  test('--llms markdown includes argument tables', async () => {
-    const { output } = await serve(createApp(), ['project', '--llms'])
+  test('--llms-full markdown includes argument tables', async () => {
+    const { output } = await serve(createApp(), ['project', '--llms-full'])
     expect(output).toContain('Arguments')
     expect(output).toContain('`id`')
   })
 
-  test('--llms markdown includes options tables', async () => {
-    const { output } = await serve(createApp(), ['project', '--llms'])
+  test('--llms-full markdown includes options tables', async () => {
+    const { output } = await serve(createApp(), ['project', '--llms-full'])
     expect(output).toContain('Options')
     expect(output).toContain('`--limit`')
   })
 
-  test('--llms json includes examples on commands', async () => {
-    const { output } = await serve(createApp(), ['project', 'deploy', '--llms', '--format', 'json'])
+  test('--llms-full json includes examples on commands', async () => {
+    const { output } = await serve(createApp(), ['project', 'deploy', '--llms-full', '--format', 'json'])
     const deployCreate = json(output).commands.find((c: any) => c.name === 'project deploy create')
     expect(deployCreate.examples).toMatchInlineSnapshot(`
       [
@@ -1269,27 +1269,268 @@ describe('--llms', () => {
     `)
   })
 
-  test('--llms json omits examples when not defined', async () => {
-    const { output } = await serve(createApp(), ['--llms', '--format', 'json'])
+  test('--llms-full json omits examples when not defined', async () => {
+    const { output } = await serve(createApp(), ['--llms-full', '--format', 'json'])
     const ping = json(output).commands.find((c: any) => c.name === 'ping')
     expect(ping.examples).toBeUndefined()
   })
 
-  test('--llms markdown includes examples section', async () => {
-    const { output } = await serve(createApp(), ['--llms'])
+  test('--llms-full markdown includes examples section', async () => {
+    const { output } = await serve(createApp(), ['--llms-full'])
     expect(output).toContain('Examples')
     expect(output).toContain('Deploy staging from main')
     expect(output).toContain('app project deploy create staging')
   })
 
-  test('--llms markdown includes output tables', async () => {
-    const { output } = await serve(createApp(), ['project', '--llms'])
+  test('--llms-full markdown includes output tables', async () => {
+    const { output } = await serve(createApp(), ['project', '--llms-full'])
     expect(output).toContain('Output')
   })
 
+  test('--llms-full --format yaml', async () => {
+    const { output } = await serve(createApp(), ['--llms-full', '--format', 'yaml'])
+    expect(output).toContain('version: incur.v1')
+  })
+})
+
+describe('--llms', () => {
+  test('outputs compact markdown table with all commands', async () => {
+    const { output } = await serve(createApp(), ['--llms'])
+    expect(output).toMatchInlineSnapshot(`
+      "# app
+
+      A comprehensive CLI application for testing.
+
+      | Command | Description |
+      |---------|-------------|
+      | \`app api\` | Proxy to HTTP API |
+      | \`app auth login\` | Log in to the service |
+      | \`app auth logout\` | Log out of the service |
+      | \`app auth status\` | Show authentication status |
+      | \`app config [key]\` | Show current configuration |
+      | \`app echo <message> [repeat]\` | Echo back arguments |
+      | \`app explode\` | Always fails |
+      | \`app explode-clac\` | Fails with IncurError |
+      | \`app noop\` | Returns nothing |
+      | \`app ping\` | Health check |
+      | \`app project create <name>\` | Create a new project |
+      | \`app project delete <id>\` | Delete a project |
+      | \`app project deploy create <env>\` | Create a deployment |
+      | \`app project deploy rollback <deployId>\` | Rollback a deployment |
+      | \`app project deploy status <deployId>\` | Check deployment status |
+      | \`app project get <id>\` | Get a project by ID |
+      | \`app project list\` | List projects |
+      | \`app slow\` | Async command |
+      | \`app stream\` | Stream chunks |
+      | \`app stream-error\` | Stream with mid-stream error |
+      | \`app stream-ok\` | Stream with ok() return |
+      | \`app stream-text\` | Stream plain text |
+      | \`app stream-throw\` | Stream that throws |
+      | \`app validate-fail <email> <age>\` | Fails validation |
+
+      Run \`app --llms-full\` for full manifest. Run \`app <command> --schema\` for argument details.
+      "
+    `)
+  })
+
+  test('json manifest has name + description only, no schema/examples', async () => {
+    const { output } = await serve(createApp(), ['--llms', '--format', 'json'])
+    expect(json(output)).toMatchInlineSnapshot(`
+      {
+        "commands": [
+          {
+            "description": "Proxy to HTTP API",
+            "name": "api",
+          },
+          {
+            "description": "Log in to the service",
+            "name": "auth login",
+          },
+          {
+            "description": "Log out of the service",
+            "name": "auth logout",
+          },
+          {
+            "description": "Show authentication status",
+            "name": "auth status",
+          },
+          {
+            "description": "Show current configuration",
+            "name": "config",
+          },
+          {
+            "description": "Echo back arguments",
+            "name": "echo",
+          },
+          {
+            "description": "Always fails",
+            "name": "explode",
+          },
+          {
+            "description": "Fails with IncurError",
+            "name": "explode-clac",
+          },
+          {
+            "description": "Returns nothing",
+            "name": "noop",
+          },
+          {
+            "description": "Health check",
+            "name": "ping",
+          },
+          {
+            "description": "Create a new project",
+            "name": "project create",
+          },
+          {
+            "description": "Delete a project",
+            "name": "project delete",
+          },
+          {
+            "description": "Create a deployment",
+            "name": "project deploy create",
+          },
+          {
+            "description": "Rollback a deployment",
+            "name": "project deploy rollback",
+          },
+          {
+            "description": "Check deployment status",
+            "name": "project deploy status",
+          },
+          {
+            "description": "Get a project by ID",
+            "name": "project get",
+          },
+          {
+            "description": "List projects",
+            "name": "project list",
+          },
+          {
+            "description": "Async command",
+            "name": "slow",
+          },
+          {
+            "description": "Stream chunks",
+            "name": "stream",
+          },
+          {
+            "description": "Stream with mid-stream error",
+            "name": "stream-error",
+          },
+          {
+            "description": "Stream with ok() return",
+            "name": "stream-ok",
+          },
+          {
+            "description": "Stream plain text",
+            "name": "stream-text",
+          },
+          {
+            "description": "Stream that throws",
+            "name": "stream-throw",
+          },
+          {
+            "description": "Fails validation",
+            "name": "validate-fail",
+          },
+        ],
+        "version": "incur.v1",
+      }
+    `)
+  })
+
+  test('scoped to group', async () => {
+    const { output } = await serve(createApp(), ['auth', '--llms'])
+    expect(output).toMatchInlineSnapshot(`
+      "# app auth
+
+      Authentication commands
+
+      | Command | Description |
+      |---------|-------------|
+      | \`app auth auth login\` | Log in to the service |
+      | \`app auth auth logout\` | Log out of the service |
+      | \`app auth auth status\` | Show authentication status |
+
+      Run \`app auth --llms-full\` for full manifest. Run \`app auth <command> --schema\` for argument details.
+      "
+    `)
+  })
+
+  test('scoped to nested group', async () => {
+    const { output } = await serve(createApp(), ['project', 'deploy', '--llms'])
+    expect(output).toMatchInlineSnapshot(`
+      "# app project deploy
+
+      Deployment commands
+
+      | Command | Description |
+      |---------|-------------|
+      | \`app project deploy project deploy create <env>\` | Create a deployment |
+      | \`app project deploy project deploy rollback <deployId>\` | Rollback a deployment |
+      | \`app project deploy project deploy status <deployId>\` | Check deployment status |
+
+      Run \`app project deploy --llms-full\` for full manifest. Run \`app project deploy <command> --schema\` for argument details.
+      "
+    `)
+  })
+
   test('--llms --format yaml', async () => {
     const { output } = await serve(createApp(), ['--llms', '--format', 'yaml'])
-    expect(output).toContain('version: incur.v1')
+    expect(output).toMatchInlineSnapshot(`
+      "version: incur.v1
+      commands:
+        - name: api
+          description: Proxy to HTTP API
+        - name: auth login
+          description: Log in to the service
+        - name: auth logout
+          description: Log out of the service
+        - name: auth status
+          description: Show authentication status
+        - name: config
+          description: Show current configuration
+        - name: echo
+          description: Echo back arguments
+        - name: explode
+          description: Always fails
+        - name: explode-clac
+          description: Fails with IncurError
+        - name: noop
+          description: Returns nothing
+        - name: ping
+          description: Health check
+        - name: project create
+          description: Create a new project
+        - name: project delete
+          description: Delete a project
+        - name: project deploy create
+          description: Create a deployment
+        - name: project deploy rollback
+          description: Rollback a deployment
+        - name: project deploy status
+          description: Check deployment status
+        - name: project get
+          description: Get a project by ID
+        - name: project list
+          description: List projects
+        - name: slow
+          description: Async command
+        - name: stream
+          description: Stream chunks
+        - name: stream-error
+          description: Stream with mid-stream error
+        - name: stream-ok
+          description: Stream with ok() return
+        - name: stream-text
+          description: Stream plain text
+        - name: stream-throw
+          description: Stream that throws
+        - name: validate-fail
+          description: Fails validation
+      "
+    `)
   })
 })
 
@@ -1461,7 +1702,7 @@ describe('root command with subcommands', () => {
         --filter-output <keys>              Filter output by key paths (e.g. foo,bar.baz,a[0,3])
         --format <toon|json|yaml|md|jsonl>  Output format
         --help                              Show help
-        --llms                              Print LLM-readable manifest
+        --llms, --llms-full                 Print LLM-readable manifest
         --mcp                               Start as MCP stdio server
         --schema                            Show JSON Schema for a command
         --token-count                       Print token count of output (instead of output)
@@ -1638,7 +1879,7 @@ describe('env', () => {
         --filter-output <keys>              Filter output by key paths (e.g. foo,bar.baz,a[0,3])
         --format <toon|json|yaml|md|jsonl>  Output format
         --help                              Show help
-        --llms                              Print LLM-readable manifest
+        --llms, --llms-full                 Print LLM-readable manifest
         --schema                            Show JSON Schema for a command
         --token-count                       Print token count of output (instead of output)
         --token-limit <n>                   Limit output to n tokens
@@ -1652,8 +1893,8 @@ describe('env', () => {
     `)
   })
 
-  test('--llms json includes schema.env', async () => {
-    const { output } = await serve(createApp(), ['auth', '--llms', '--format', 'json'])
+  test('--llms-full json includes schema.env', async () => {
+    const { output } = await serve(createApp(), ['auth', '--llms-full', '--format', 'json'])
     const login = json(output).commands.find((c: any) => c.name === 'auth login')
     expect(login.schema.env.properties).toMatchInlineSnapshot(`
       {
@@ -1670,8 +1911,8 @@ describe('env', () => {
     `)
   })
 
-  test('--llms markdown includes env vars table', async () => {
-    const { output } = await serve(createApp(), ['auth', '--llms'])
+  test('--llms-full markdown includes env vars table', async () => {
+    const { output } = await serve(createApp(), ['auth', '--llms-full'])
     expect(output).toContain('Environment Variables')
     expect(output).toContain('`AUTH_TOKEN`')
     expect(output).toContain('`AUTH_HOST`')
@@ -2631,6 +2872,7 @@ describe('.well-known/skills', () => {
       "---
       name: app-ping
       description: Health check. Run \`app ping --help\` for usage details.
+      requires_bin: app
       command: app ping
       ---