Merge pull request #1097 from NousResearch/hermes/hermes-c877bdeb

feat: native Anthropic provider with Claude Code credential auto-discovery

agent/anthropic_adapter.py

@@ -0,0 +1,466 @@
+"""Anthropic Messages API adapter for Hermes Agent.
+
+Translates between Hermes's internal OpenAI-style message format and
+Anthropic's Messages API. Follows the same pattern as the codex_responses
+adapter — all provider-specific logic is isolated here.
+
+Auth supports:
+  - Regular API keys (sk-ant-api*) → x-api-key header
+  - OAuth setup-tokens (sk-ant-oat*) → Bearer auth + beta header
+  - Claude Code credentials (~/.claude.json or ~/.claude/.credentials.json) → Bearer auth
+"""
+
+import json
+import logging
+import os
+from pathlib import Path
+from types import SimpleNamespace
+from typing import Any, Dict, List, Optional, Tuple
+
+try:
+    import anthropic as _anthropic_sdk
+except ImportError:
+    _anthropic_sdk = None  # type: ignore[assignment]
+
+logger = logging.getLogger(__name__)
+
+THINKING_BUDGET = {"xhigh": 32000, "high": 16000, "medium": 8000, "low": 4000}
+
+# Beta headers for enhanced features (sent with ALL auth types)
+_COMMON_BETAS = [
+    "interleaved-thinking-2025-05-14",
+    "fine-grained-tool-streaming-2025-05-14",
+]
+
+# Additional beta headers required for OAuth/subscription auth
+_OAUTH_ONLY_BETAS = [
+    "oauth-2025-04-20",
+]
+
+
+def _is_oauth_token(key: str) -> bool:
+    """Check if the key is an OAuth/setup token (not a regular Console API key).
+
+    Regular API keys start with 'sk-ant-api'. Everything else (setup-tokens
+    starting with 'sk-ant-oat', managed keys, JWTs, etc.) needs Bearer auth.
+    """
+    if not key:
+        return False
+    # Regular Console API keys use x-api-key header
+    if key.startswith("sk-ant-api"):
+        return False
+    # Everything else (setup-tokens, managed keys, JWTs) uses Bearer auth
+    return True
+
+
+def build_anthropic_client(api_key: str, base_url: str = None):
+    """Create an Anthropic client, auto-detecting setup-tokens vs API keys.
+
+    Returns an anthropic.Anthropic instance.
+    """
+    if _anthropic_sdk is None:
+        raise ImportError(
+            "The 'anthropic' package is required for the Anthropic provider. "
+            "Install it with: pip install 'anthropic>=0.39.0'"
+        )
+    from httpx import Timeout
+
+    kwargs = {
+        "timeout": Timeout(timeout=900.0, connect=10.0),
+    }
+    if base_url:
+        kwargs["base_url"] = base_url
+
+    if _is_oauth_token(api_key):
+        # OAuth access token / setup-token → Bearer auth + beta headers
+        all_betas = _COMMON_BETAS + _OAUTH_ONLY_BETAS
+        kwargs["auth_token"] = api_key
+        kwargs["default_headers"] = {"anthropic-beta": ",".join(all_betas)}
+    else:
+        # Regular API key → x-api-key header + common betas
+        kwargs["api_key"] = api_key
+        if _COMMON_BETAS:
+            kwargs["default_headers"] = {"anthropic-beta": ",".join(_COMMON_BETAS)}
+
+    return _anthropic_sdk.Anthropic(**kwargs)
+
+
+def read_claude_code_credentials() -> Optional[Dict[str, Any]]:
+    """Read credentials from Claude Code's config files.
+
+    Checks two locations (in order):
+      1. ~/.claude.json — top-level primaryApiKey (native binary, v2.x)
+      2. ~/.claude/.credentials.json — claudeAiOauth block (npm/legacy installs)
+
+    Returns dict with {accessToken, refreshToken?, expiresAt?} or None.
+    """
+    # 1. Native binary (v2.x): ~/.claude.json with top-level primaryApiKey
+    claude_json = Path.home() / ".claude.json"
+    if claude_json.exists():
+        try:
+            data = json.loads(claude_json.read_text(encoding="utf-8"))
+            primary_key = data.get("primaryApiKey", "")
+            if primary_key:
+                return {
+                    "accessToken": primary_key,
+                    "refreshToken": "",
+                    "expiresAt": 0,  # Managed keys don't have a user-visible expiry
+                }
+        except (json.JSONDecodeError, OSError, IOError) as e:
+            logger.debug("Failed to read ~/.claude.json: %s", e)
+
+    # 2. Legacy/npm installs: ~/.claude/.credentials.json
+    cred_path = Path.home() / ".claude" / ".credentials.json"
+    if cred_path.exists():
+        try:
+            data = json.loads(cred_path.read_text(encoding="utf-8"))
+            oauth_data = data.get("claudeAiOauth")
+            if oauth_data and isinstance(oauth_data, dict):
+                access_token = oauth_data.get("accessToken", "")
+                if access_token:
+                    return {
+                        "accessToken": access_token,
+                        "refreshToken": oauth_data.get("refreshToken", ""),
+                        "expiresAt": oauth_data.get("expiresAt", 0),
+                    }
+        except (json.JSONDecodeError, OSError, IOError) as e:
+            logger.debug("Failed to read ~/.claude/.credentials.json: %s", e)
+
+    return None
+
+
+def is_claude_code_token_valid(creds: Dict[str, Any]) -> bool:
+    """Check if Claude Code credentials have a non-expired access token."""
+    import time
+
+    expires_at = creds.get("expiresAt", 0)
+    if not expires_at:
+        # No expiry set (managed keys) — valid if token is present
+        return bool(creds.get("accessToken"))
+
+    # expiresAt is in milliseconds since epoch
+    now_ms = int(time.time() * 1000)
+    # Allow 60 seconds of buffer
+    return now_ms < (expires_at - 60_000)
+
+
+def resolve_anthropic_token() -> Optional[str]:
+    """Resolve an Anthropic token from all available sources.
+
+    Priority:
+      1. ANTHROPIC_API_KEY env var (regular API key)
+      2. ANTHROPIC_TOKEN env var (OAuth/setup token)
+      3. Claude Code credentials (~/.claude.json or ~/.claude/.credentials.json)
+
+    Returns the token string or None.
+    """
+    # 1. Regular API key
+    api_key = os.getenv("ANTHROPIC_API_KEY", "").strip()
+    if api_key:
+        return api_key
+
+    # 2. OAuth/setup token env var
+    token = os.getenv("ANTHROPIC_TOKEN", "").strip()
+    if token:
+        return token
+
+    # Also check CLAUDE_CODE_OAUTH_TOKEN (used by Claude Code for setup-tokens)
+    cc_token = os.getenv("CLAUDE_CODE_OAUTH_TOKEN", "").strip()
+    if cc_token:
+        return cc_token
+
+    # 3. Claude Code credential file
+    creds = read_claude_code_credentials()
+    if creds and is_claude_code_token_valid(creds):
+        logger.debug("Using Claude Code credentials (auto-detected)")
+        return creds["accessToken"]
+    elif creds:
+        logger.debug("Claude Code credentials expired — run 'claude' to refresh")
+
+    return None
+
+
+# ---------------------------------------------------------------------------
+# Message / tool / response format conversion
+# ---------------------------------------------------------------------------
+
+
+def normalize_model_name(model: str) -> str:
+    """Normalize a model name for the Anthropic API.
+
+    - Strips 'anthropic/' prefix (OpenRouter format, case-insensitive)
+    """
+    lower = model.lower()
+    if lower.startswith("anthropic/"):
+        model = model[len("anthropic/"):]
+    return model
+
+
+def _sanitize_tool_id(tool_id: str) -> str:
+    """Sanitize a tool call ID for the Anthropic API.
+
+    Anthropic requires IDs matching [a-zA-Z0-9_-]. Replace invalid
+    characters with underscores and ensure non-empty.
+    """
+    import re
+    if not tool_id:
+        return "tool_0"
+    sanitized = re.sub(r"[^a-zA-Z0-9_-]", "_", tool_id)
+    return sanitized or "tool_0"
+
+
+def convert_tools_to_anthropic(tools: List[Dict]) -> List[Dict]:
+    """Convert OpenAI tool definitions to Anthropic format."""
+    if not tools:
+        return []
+    result = []
+    for t in tools:
+        fn = t.get("function", {})
+        result.append({
+            "name": fn.get("name", ""),
+            "description": fn.get("description", ""),
+            "input_schema": fn.get("parameters", {"type": "object", "properties": {}}),
+        })
+    return result
+
+
+def convert_messages_to_anthropic(
+    messages: List[Dict],
+) -> Tuple[Optional[Any], List[Dict]]:
+    """Convert OpenAI-format messages to Anthropic format.
+
+    Returns (system_prompt, anthropic_messages).
+    System messages are extracted since Anthropic takes them as a separate param.
+    system_prompt is a string or list of content blocks (when cache_control present).
+    """
+    system = None
+    result = []
+
+    for m in messages:
+        role = m.get("role", "user")
+        content = m.get("content", "")
+
+        if role == "system":
+            if isinstance(content, list):
+                # Preserve cache_control markers on content blocks
+                has_cache = any(
+                    p.get("cache_control") for p in content if isinstance(p, dict)
+                )
+                if has_cache:
+                    system = [p for p in content if isinstance(p, dict)]
+                else:
+                    system = "\n".join(
+                        p["text"] for p in content if p.get("type") == "text"
+                    )
+            else:
+                system = content
+            continue
+
+        if role == "assistant":
+            blocks = []
+            if content:
+                text = content if isinstance(content, str) else json.dumps(content)
+                blocks.append({"type": "text", "text": text})
+            for tc in m.get("tool_calls", []):
+                fn = tc.get("function", {})
+                args = fn.get("arguments", "{}")
+                try:
+                    parsed_args = json.loads(args) if isinstance(args, str) else args
+                except (json.JSONDecodeError, ValueError):
+                    parsed_args = {}
+                blocks.append({
+                    "type": "tool_use",
+                    "id": _sanitize_tool_id(tc.get("id", "")),
+                    "name": fn.get("name", ""),
+                    "input": parsed_args,
+                })
+            # Anthropic rejects empty assistant content
+            effective = blocks or content
+            if not effective or effective == "":
+                effective = [{"type": "text", "text": "(empty)"}]
+            result.append({"role": "assistant", "content": effective})
+            continue
+
+        if role == "tool":
+            # Sanitize tool_use_id and ensure non-empty content
+            result_content = content if isinstance(content, str) else json.dumps(content)
+            if not result_content:
+                result_content = "(no output)"
+            tool_result = {
+                "type": "tool_result",
+                "tool_use_id": _sanitize_tool_id(m.get("tool_call_id", "")),
+                "content": result_content,
+            }
+            # Merge consecutive tool results into one user message
+            if (
+                result
+                and result[-1]["role"] == "user"
+                and isinstance(result[-1]["content"], list)
+                and result[-1]["content"]
+                and result[-1]["content"][0].get("type") == "tool_result"
+            ):
+                result[-1]["content"].append(tool_result)
+            else:
+                result.append({"role": "user", "content": [tool_result]})
+            continue
+
+        # Regular user message
+        result.append({"role": "user", "content": content})
+
+    # Strip orphaned tool_use blocks (no matching tool_result follows)
+    tool_result_ids = set()
+    for m in result:
+        if m["role"] == "user" and isinstance(m["content"], list):
+            for block in m["content"]:
+                if block.get("type") == "tool_result":
+                    tool_result_ids.add(block.get("tool_use_id"))
+    for m in result:
+        if m["role"] == "assistant" and isinstance(m["content"], list):
+            m["content"] = [
+                b
+                for b in m["content"]
+                if b.get("type") != "tool_use" or b.get("id") in tool_result_ids
+            ]
+            if not m["content"]:
+                m["content"] = [{"type": "text", "text": "(tool call removed)"}]
+
+    # Enforce strict role alternation (Anthropic rejects consecutive same-role messages)
+    fixed = []
+    for m in result:
+        if fixed and fixed[-1]["role"] == m["role"]:
+            if m["role"] == "user":
+                # Merge consecutive user messages
+                prev_content = fixed[-1]["content"]
+                curr_content = m["content"]
+                if isinstance(prev_content, str) and isinstance(curr_content, str):
+                    fixed[-1]["content"] = prev_content + "\n" + curr_content
+                elif isinstance(prev_content, list) and isinstance(curr_content, list):
+                    fixed[-1]["content"] = prev_content + curr_content
+                else:
+                    # Mixed types — wrap string in list
+                    if isinstance(prev_content, str):
+                        prev_content = [{"type": "text", "text": prev_content}]
+                    if isinstance(curr_content, str):
+                        curr_content = [{"type": "text", "text": curr_content}]
+                    fixed[-1]["content"] = prev_content + curr_content
+            else:
+                # Consecutive assistant messages — merge text content
+                prev_blocks = fixed[-1]["content"]
+                curr_blocks = m["content"]
+                if isinstance(prev_blocks, list) and isinstance(curr_blocks, list):
+                    fixed[-1]["content"] = prev_blocks + curr_blocks
+                elif isinstance(prev_blocks, str) and isinstance(curr_blocks, str):
+                    fixed[-1]["content"] = prev_blocks + "\n" + curr_blocks
+                else:
+                    # Keep the later message
+                    fixed[-1] = m
+        else:
+            fixed.append(m)
+    result = fixed
+
+    return system, result
+
+
+def build_anthropic_kwargs(
+    model: str,
+    messages: List[Dict],
+    tools: Optional[List[Dict]],
+    max_tokens: Optional[int],
+    reasoning_config: Optional[Dict[str, Any]],
+    tool_choice: Optional[str] = None,
+) -> Dict[str, Any]:
+    """Build kwargs for anthropic.messages.create()."""
+    system, anthropic_messages = convert_messages_to_anthropic(messages)
+    anthropic_tools = convert_tools_to_anthropic(tools) if tools else []
+
+    model = normalize_model_name(model)
+    effective_max_tokens = max_tokens or 16384
+
+    kwargs: Dict[str, Any] = {
+        "model": model,
+        "messages": anthropic_messages,
+        "max_tokens": effective_max_tokens,
+    }
+
+    if system:
+        kwargs["system"] = system
+
+    if anthropic_tools:
+        kwargs["tools"] = anthropic_tools
+        # Map OpenAI tool_choice to Anthropic format
+        if tool_choice == "auto" or tool_choice is None:
+            kwargs["tool_choice"] = {"type": "auto"}
+        elif tool_choice == "required":
+            kwargs["tool_choice"] = {"type": "any"}
+        elif tool_choice == "none":
+            pass  # Don't send tool_choice — Anthropic will use tools if needed
+        elif isinstance(tool_choice, str):
+            # Specific tool name
+            kwargs["tool_choice"] = {"type": "tool", "name": tool_choice}
+
+    # Map reasoning_config to Anthropic's thinking parameter
+    # Newer models (4.6+) prefer "adaptive" thinking; older models use "enabled"
+    if reasoning_config and isinstance(reasoning_config, dict):
+        if reasoning_config.get("enabled") is not False:
+            effort = reasoning_config.get("effort", "medium")
+            budget = THINKING_BUDGET.get(effort, 8000)
+            # Use adaptive thinking for 4.5+ models (they deprecate type=enabled)
+            if any(v in model for v in ("4-6", "4-5", "4.6", "4.5")):
+                kwargs["thinking"] = {"type": "adaptive", "budget_tokens": budget}
+            else:
+                kwargs["thinking"] = {"type": "enabled", "budget_tokens": budget}
+                # Anthropic requires temperature=1 when thinking is enabled on older models
+                kwargs["temperature"] = 1
+            kwargs["max_tokens"] = max(effective_max_tokens, budget + 4096)
+
+    return kwargs
+
+
+def normalize_anthropic_response(
+    response,
+) -> Tuple[SimpleNamespace, str]:
+    """Normalize Anthropic response to match the shape expected by AIAgent.
+
+    Returns (assistant_message, finish_reason) where assistant_message has
+    .content, .tool_calls, and .reasoning attributes.
+    """
+    text_parts = []
+    reasoning_parts = []
+    tool_calls = []
+
+    for block in response.content:
+        if block.type == "text":
+            text_parts.append(block.text)
+        elif block.type == "thinking":
+            reasoning_parts.append(block.thinking)
+        elif block.type == "tool_use":
+            tool_calls.append(
+                SimpleNamespace(
+                    id=block.id,
+                    type="function",
+                    function=SimpleNamespace(
+                        name=block.name,
+                        arguments=json.dumps(block.input),
+                    ),
+                )
+            )
+
+    # Map Anthropic stop_reason to OpenAI finish_reason
+    stop_reason_map = {
+        "end_turn": "stop",
+        "tool_use": "tool_calls",
+        "max_tokens": "length",
+        "stop_sequence": "stop",
+    }
+    finish_reason = stop_reason_map.get(response.stop_reason, "stop")
+
+    return (
+        SimpleNamespace(
+            content="\n".join(text_parts) if text_parts else None,
+            tool_calls=tool_calls or None,
+            reasoning="\n\n".join(reasoning_parts) if reasoning_parts else None,
+            reasoning_content=None,
+            reasoning_details=None,
+        ),
+        finish_reason,
+    )

agent/auxiliary_client.py

@@ -51,6 +51,7 @@ _API_KEY_PROVIDER_AUX_MODELS: Dict[str, str] = {
     "kimi-coding": "kimi-k2-turbo-preview",
     "minimax": "MiniMax-M2.5-highspeed",
     "minimax-cn": "MiniMax-M2.5-highspeed",
+    "anthropic": "claude-haiku-4-5-20251001",
 }
 
 # OpenRouter app attribution headers

agent/model_metadata.py

@@ -41,6 +41,15 @@ DEFAULT_CONTEXT_LENGTHS = {
     "anthropic/claude-sonnet-4": 200000,
     "anthropic/claude-sonnet-4-20250514": 200000,
     "anthropic/claude-haiku-4.5": 200000,
+    # Bare Anthropic model IDs (for native API provider)
+    "claude-opus-4-6": 200000,
+    "claude-sonnet-4-6": 200000,
+    "claude-opus-4-5-20251101": 200000,
+    "claude-sonnet-4-5-20250929": 200000,
+    "claude-opus-4-1-20250805": 200000,
+    "claude-opus-4-20250514": 200000,
+    "claude-sonnet-4-20250514": 200000,
+    "claude-haiku-4-5-20251001": 200000,
     "openai/gpt-4o": 128000,
     "openai/gpt-4-turbo": 128000,
     "openai/gpt-4o-mini": 128000,

hermes_cli/auth.py

@@ -132,6 +132,13 @@ PROVIDER_REGISTRY: Dict[str, ProviderConfig] = {
         api_key_env_vars=("MINIMAX_API_KEY",),
         base_url_env_var="MINIMAX_BASE_URL",
     ),
+    "anthropic": ProviderConfig(
+        id="anthropic",
+        name="Anthropic",
+        auth_type="api_key",
+        inference_base_url="https://api.anthropic.com",
+        api_key_env_vars=("ANTHROPIC_API_KEY", "ANTHROPIC_TOKEN", "CLAUDE_CODE_OAUTH_TOKEN"),
+    ),
     "minimax-cn": ProviderConfig(
         id="minimax-cn",
         name="MiniMax (China)",
@@ -516,6 +523,7 @@ def resolve_provider(
         "glm": "zai", "z-ai": "zai", "z.ai": "zai", "zhipu": "zai",
         "kimi": "kimi-coding", "moonshot": "kimi-coding",
         "minimax-china": "minimax-cn", "minimax_cn": "minimax-cn",
+        "claude": "anthropic", "claude-code": "anthropic",
     }
     normalized = _PROVIDER_ALIASES.get(normalized, normalized)
 

hermes_cli/main.py

@@ -748,6 +748,7 @@ def cmd_model(args):
         "openrouter": "OpenRouter",
         "nous": "Nous Portal",
         "openai-codex": "OpenAI Codex",
+        "anthropic": "Anthropic",
         "zai": "Z.AI / GLM",
         "kimi-coding": "Kimi / Moonshot",
         "minimax": "MiniMax",
@@ -766,6 +767,7 @@ def cmd_model(args):
         ("openrouter", "OpenRouter (100+ models, pay-per-use)"),
         ("nous", "Nous Portal (Nous Research subscription)"),
         ("openai-codex", "OpenAI Codex"),
+        ("anthropic", "Anthropic (Claude models — API key or Claude Code)"),
         ("zai", "Z.AI / GLM (Zhipu AI direct API)"),
         ("kimi-coding", "Kimi / Moonshot (Moonshot AI direct API)"),
         ("minimax", "MiniMax (global direct API)"),
@@ -834,6 +836,8 @@ def cmd_model(args):
         _model_flow_named_custom(config, _custom_provider_map[selected_provider])
     elif selected_provider == "remove-custom":
         _remove_custom_provider(config)
+    elif selected_provider == "anthropic":
+        _model_flow_anthropic(config, current_model)
     elif selected_provider == "kimi-coding":
         _model_flow_kimi(config, current_model)
     elif selected_provider in ("zai", "minimax", "minimax-cn"):
@@ -1570,6 +1574,140 @@ def _model_flow_api_key_provider(config, provider_id, current_model=""):
         print("No change.")
 
 
+def _model_flow_anthropic(config, current_model=""):
+    """Flow for Anthropic provider — setup-token, API key, or Claude Code creds."""
+    import os
+    from hermes_cli.auth import (
+        PROVIDER_REGISTRY, _prompt_model_selection, _save_model_choice,
+        _update_config_for_provider, deactivate_provider,
+    )
+    from hermes_cli.config import get_env_value, save_env_value, load_config, save_config
+    from hermes_cli.models import _PROVIDER_MODELS
+
+    pconfig = PROVIDER_REGISTRY["anthropic"]
+
+    # Check for existing credentials
+    existing_key = (
+        get_env_value("ANTHROPIC_API_KEY")
+        or os.getenv("ANTHROPIC_API_KEY", "")
+        or get_env_value("ANTHROPIC_TOKEN")
+        or os.getenv("ANTHROPIC_TOKEN", "")
+    )
+    cc_available = False
+    try:
+        from agent.anthropic_adapter import read_claude_code_credentials, is_claude_code_token_valid
+        cc_creds = read_claude_code_credentials()
+        if cc_creds and is_claude_code_token_valid(cc_creds):
+            cc_available = True
+    except Exception:
+        pass
+
+    if existing_key:
+        print(f"  Anthropic credentials: {existing_key[:12]}... ✓")
+        print()
+        try:
+            update = input("Update credentials? [y/N]: ").strip().lower()
+        except (KeyboardInterrupt, EOFError):
+            update = ""
+        if update != "y":
+            pass  # skip to model selection
+        else:
+            existing_key = ""  # fall through to auth choice below
+    elif cc_available:
+        print("  Claude Code credentials: ✓ (auto-detected)")
+        print()
+    
+    if not existing_key and not cc_available:
+        # No credentials — show auth method choice
+        print()
+        print("  Choose authentication method:")
+        print()
+        print("    1. Claude Pro/Max subscription (setup-token)")
+        print("    2. Anthropic API key (pay-per-token)")
+        print("    3. Cancel")
+        print()
+        try:
+            choice = input("  Choice [1/2/3]: ").strip()
+        except (KeyboardInterrupt, EOFError):
+            print()
+            return
+
+        if choice == "1":
+            print()
+            print("  To get a setup-token from your Claude subscription:")
+            print()
+            print("    1. Install Claude Code:  npm install -g @anthropic-ai/claude-code")
+            print("    2. Run:                  claude setup-token")
+            print("    3. Open the URL it prints in your browser")
+            print("    4. Log in and click \"Authorize\"")
+            print("    5. Paste the auth code back into Claude Code")
+            print("    6. Copy the resulting sk-ant-oat01-... token")
+            print()
+            try:
+                token = input("  Paste setup-token here: ").strip()
+            except (KeyboardInterrupt, EOFError):
+                print()
+                return
+            if not token:
+                print("  Cancelled.")
+                return
+            save_env_value("ANTHROPIC_API_KEY", token)
+            print("  ✓ Setup-token saved.")
+
+        elif choice == "2":
+            print()
+            print("  Get an API key at: https://console.anthropic.com/settings/keys")
+            print()
+            try:
+                api_key = input("  API key (sk-ant-api03-...): ").strip()
+            except (KeyboardInterrupt, EOFError):
+                print()
+                return
+            if not api_key:
+                print("  Cancelled.")
+                return
+            save_env_value("ANTHROPIC_API_KEY", api_key)
+            print("  ✓ API key saved.")
+
+        else:
+            print("  No change.")
+            return
+    print()
+
+    # Model selection
+    model_list = _PROVIDER_MODELS.get("anthropic", [])
+    if model_list:
+        selected = _prompt_model_selection(model_list, current_model=current_model)
+    else:
+        try:
+            selected = input("Model name (e.g., claude-sonnet-4-20250514): ").strip()
+        except (KeyboardInterrupt, EOFError):
+            selected = None
+
+    if selected:
+        # Clear custom endpoint if set
+        if get_env_value("OPENAI_BASE_URL"):
+            save_env_value("OPENAI_BASE_URL", "")
+            save_env_value("OPENAI_API_KEY", "")
+
+        _save_model_choice(selected)
+
+        # Update config with provider
+        cfg = load_config()
+        model = cfg.get("model")
+        if not isinstance(model, dict):
+            model = {"default": model} if model else {}
+            cfg["model"] = model
+        model["provider"] = "anthropic"
+        model["base_url"] = pconfig.inference_base_url
+        save_config(cfg)
+        deactivate_provider()
+
+        print(f"Default model set to: {selected} (via Anthropic)")
+    else:
+        print("No change.")
+
+
 def cmd_login(args):
     """Authenticate Hermes CLI with a provider."""
     from hermes_cli.auth import login_command
@@ -2050,7 +2188,7 @@ For more help on a command:
     )
     chat_parser.add_argument(
         "--provider",
-        choices=["auto", "openrouter", "nous", "openai-codex", "zai", "kimi-coding", "minimax", "minimax-cn"],
+        choices=["auto", "openrouter", "nous", "openai-codex", "anthropic", "zai", "kimi-coding", "minimax", "minimax-cn"],
         default=None,
         help="Inference provider (default: auto)"
     )

hermes_cli/models.py

@@ -68,6 +68,15 @@ _PROVIDER_MODELS: dict[str, list[str]] = {
         "MiniMax-M2.5-highspeed",
         "MiniMax-M2.1",
     ],
+    "anthropic": [
+        "claude-opus-4-6",
+        "claude-sonnet-4-6",
+        "claude-opus-4-5-20251101",
+        "claude-sonnet-4-5-20250929",
+        "claude-opus-4-20250514",
+        "claude-sonnet-4-20250514",
+        "claude-haiku-4-5-20251001",
+    ],
 }
 
 _PROVIDER_LABELS = {
@@ -78,6 +87,7 @@ _PROVIDER_LABELS = {
     "kimi-coding": "Kimi / Moonshot",
     "minimax": "MiniMax",
     "minimax-cn": "MiniMax (China)",
+    "anthropic": "Anthropic",
     "custom": "Custom endpoint",
 }
 
@@ -90,6 +100,8 @@ _PROVIDER_ALIASES = {
     "moonshot": "kimi-coding",
     "minimax-china": "minimax-cn",
     "minimax_cn": "minimax-cn",
+    "claude": "anthropic",
+    "claude-code": "anthropic",
 }
 
 
@@ -123,7 +135,7 @@ def list_available_providers() -> list[dict[str, str]]:
     # Canonical providers in display order
     _PROVIDER_ORDER = [
         "openrouter", "nous", "openai-codex",
-        "zai", "kimi-coding", "minimax", "minimax-cn",
+        "zai", "kimi-coding", "minimax", "minimax-cn", "anthropic",
     ]
     # Build reverse alias map
     aliases_for: dict[str, list[str]] = {}
@@ -234,9 +246,56 @@ def provider_model_ids(provider: Optional[str]) -> list[str]:
                     return live
         except Exception:
             pass
+    if normalized == "anthropic":
+        live = _fetch_anthropic_models()
+        if live:
+            return live
     return list(_PROVIDER_MODELS.get(normalized, []))
 
 
+def _fetch_anthropic_models(timeout: float = 5.0) -> Optional[list[str]]:
+    """Fetch available models from the Anthropic /v1/models endpoint.
+
+    Uses resolve_anthropic_token() to find credentials (env vars or
+    Claude Code auto-discovery).  Returns sorted model IDs or None.
+    """
+    try:
+        from agent.anthropic_adapter import resolve_anthropic_token, _is_oauth_token
+    except ImportError:
+        return None
+
+    token = resolve_anthropic_token()
+    if not token:
+        return None
+
+    headers: dict[str, str] = {"anthropic-version": "2023-06-01"}
+    if _is_oauth_token(token):
+        headers["Authorization"] = f"Bearer {token}"
+        headers["anthropic-beta"] = "oauth-2025-04-20"
+    else:
+        headers["x-api-key"] = token
+
+    req = urllib.request.Request(
+        "https://api.anthropic.com/v1/models",
+        headers=headers,
+    )
+    try:
+        with urllib.request.urlopen(req, timeout=timeout) as resp:
+            data = json.loads(resp.read().decode())
+            models = [m["id"] for m in data.get("data", []) if m.get("id")]
+            # Sort: latest/largest first (opus > sonnet > haiku, higher version first)
+            return sorted(models, key=lambda m: (
+                "opus" not in m,      # opus first
+                "sonnet" not in m,    # then sonnet
+                "haiku" not in m,     # then haiku
+                m,                    # alphabetical within tier
+            ))
+    except Exception as e:
+        import logging
+        logging.getLogger(__name__).debug("Failed to fetch Anthropic models: %s", e)
+        return None
+
+
 def fetch_api_models(
     api_key: Optional[str],
     base_url: Optional[str],

hermes_cli/runtime_provider.py

@@ -153,6 +153,24 @@ def resolve_runtime_provider(
             "requested_provider": requested_provider,
         }
 
+    # Anthropic (native Messages API)
+    if provider == "anthropic":
+        from agent.anthropic_adapter import resolve_anthropic_token
+        token = resolve_anthropic_token()
+        if not token:
+            raise AuthError(
+                "No Anthropic credentials found. Set ANTHROPIC_API_KEY, "
+                "run 'claude setup-token', or authenticate with 'claude /login'."
+            )
+        return {
+            "provider": "anthropic",
+            "api_mode": "anthropic_messages",
+            "base_url": "https://api.anthropic.com",
+            "api_key": token,
+            "source": "env",
+            "requested_provider": requested_provider,
+        }
+
     # API-key providers (z.ai/GLM, Kimi, MiniMax, MiniMax-CN)
     pconfig = PROVIDER_REGISTRY.get(provider)
     if pconfig and pconfig.auth_type == "api_key":

hermes_cli/setup.py

@@ -689,6 +689,7 @@ def setup_model_provider(config: dict):
         "Kimi / Moonshot (Kimi coding models)",
         "MiniMax (global endpoint)",
         "MiniMax China (mainland China endpoint)",
+        "Anthropic (Claude models — API key or Claude Code subscription)",
     ]
     if keep_label:
         provider_choices.append(keep_label)
@@ -1068,7 +1069,74 @@ def setup_model_provider(config: dict):
         _update_config_for_provider("minimax-cn", pconfig.inference_base_url)
         _set_model_provider(config, "minimax-cn", pconfig.inference_base_url)
 
-    # else: provider_idx == 8 (Keep current) — only shown when a provider already exists
+    elif provider_idx == 8:  # Anthropic
+        selected_provider = "anthropic"
+        print()
+        print_header("Anthropic Authentication")
+        from hermes_cli.auth import PROVIDER_REGISTRY
+        pconfig = PROVIDER_REGISTRY["anthropic"]
+
+        # Check for Claude Code credential auto-discovery
+        from agent.anthropic_adapter import read_claude_code_credentials, is_claude_code_token_valid
+        cc_creds = read_claude_code_credentials()
+        if cc_creds and is_claude_code_token_valid(cc_creds):
+            print_success("Found valid Claude Code credentials (~/.claude/.credentials.json)")
+            if prompt_yes_no("Use these credentials?", True):
+                print_success("Using Claude Code subscription credentials")
+            else:
+                cc_creds = None
+
+        existing_key = get_env_value("ANTHROPIC_API_KEY") or get_env_value("ANTHROPIC_TOKEN")
+
+        if not (cc_creds and is_claude_code_token_valid(cc_creds)):
+            if existing_key:
+                print_info(f"Current credentials: {existing_key[:12]}...")
+                if not prompt_yes_no("Update credentials?", False):
+                    # User wants to keep existing — skip auth prompt entirely
+                    existing_key = "KEEP"  # truthy sentinel to skip auth choice
+
+            if not existing_key and not (cc_creds and is_claude_code_token_valid(cc_creds)):
+                auth_choices = [
+                    "Claude Pro/Max subscription (setup-token)",
+                    "Anthropic API key (pay-per-token)",
+                ]
+                auth_idx = prompt_choice("Choose authentication method:", auth_choices, 0)
+
+                if auth_idx == 0:
+                    print()
+                    print_info("To get a setup-token from your Claude subscription:")
+                    print_info("  1. Install Claude Code:  npm install -g @anthropic-ai/claude-code")
+                    print_info("  2. Run:                  claude setup-token")
+                    print_info("  3. Open the URL it prints in your browser")
+                    print_info("  4. Log in and click \"Authorize\"")
+                    print_info("  5. Paste the auth code back into Claude Code")
+                    print_info("  6. Copy the resulting sk-ant-oat01-... token")
+                    print()
+                    token = prompt("Paste setup-token here", password=True)
+                    if token:
+                        save_env_value("ANTHROPIC_API_KEY", token)
+                        print_success("Setup-token saved")
+                    else:
+                        print_warning("Skipped — agent won't work without credentials")
+                else:
+                    print()
+                    print_info("Get an API key at: https://console.anthropic.com/settings/keys")
+                    print()
+                    api_key = prompt("API key (sk-ant-api03-...)", password=True)
+                    if api_key:
+                        save_env_value("ANTHROPIC_API_KEY", api_key)
+                        print_success("API key saved")
+                    else:
+                        print_warning("Skipped — agent won't work without credentials")
+
+        # Clear custom endpoint vars if switching
+        if existing_custom:
+            save_env_value("OPENAI_BASE_URL", "")
+            save_env_value("OPENAI_API_KEY", "")
+        _update_config_for_provider("anthropic", pconfig.inference_base_url)
+        _set_model_provider(config, "anthropic", pconfig.inference_base_url)
+
+    # else: provider_idx == 9 (Keep current) — only shown when a provider already exists
 
     # ── OpenRouter API Key for tools (if not already set) ──
     # Tools (vision, web, MoA) use OpenRouter independently of the main provider.
@@ -1081,6 +1149,7 @@ def setup_model_provider(config: dict):
         "kimi-coding",
         "minimax",
         "minimax-cn",
+        "anthropic",
     ) and not get_env_value("OPENROUTER_API_KEY"):
         print()
         print_header("OpenRouter API Key (for tools)")
@@ -1174,6 +1243,79 @@ def setup_model_provider(config: dict):
                 config, selected_provider, current_model,
                 prompt_choice, prompt,
             )
+            if is_coding_plan:
+                zai_models = ["glm-4.7", "glm-4.5", "glm-4.5-flash"]
+            else:
+                zai_models = ["glm-5", "glm-4.7", "glm-4.5", "glm-4.5-flash"]
+            model_choices = list(zai_models)
+            model_choices.append("Custom model")
+            model_choices.append(f"Keep current ({current_model})")
+
+            keep_idx = len(model_choices) - 1
+            model_idx = prompt_choice("Select default model:", model_choices, keep_idx)
+
+            if model_idx < len(zai_models):
+                _set_default_model(config, zai_models[model_idx])
+            elif model_idx == len(zai_models):
+                custom = prompt("Enter model name")
+                if custom:
+                    _set_default_model(config, custom)
+            # else: keep current
+        elif selected_provider == "kimi-coding":
+            kimi_models = ["kimi-k2.5", "kimi-k2-thinking", "kimi-k2-turbo-preview"]
+            model_choices = list(kimi_models)
+            model_choices.append("Custom model")
+            model_choices.append(f"Keep current ({current_model})")
+
+            keep_idx = len(model_choices) - 1
+            model_idx = prompt_choice("Select default model:", model_choices, keep_idx)
+
+            if model_idx < len(kimi_models):
+                _set_default_model(config, kimi_models[model_idx])
+            elif model_idx == len(kimi_models):
+                custom = prompt("Enter model name")
+                if custom:
+                    _set_default_model(config, custom)
+            # else: keep current
+        elif selected_provider in ("minimax", "minimax-cn"):
+            minimax_models = ["MiniMax-M2.5", "MiniMax-M2.5-highspeed", "MiniMax-M2.1"]
+            model_choices = list(minimax_models)
+            model_choices.append("Custom model")
+            model_choices.append(f"Keep current ({current_model})")
+
+            keep_idx = len(model_choices) - 1
+            model_idx = prompt_choice("Select default model:", model_choices, keep_idx)
+
+            if model_idx < len(minimax_models):
+                _set_default_model(config, minimax_models[model_idx])
+            elif model_idx == len(minimax_models):
+                custom = prompt("Enter model name")
+                if custom:
+                    _set_default_model(config, custom)
+            # else: keep current
+        elif selected_provider == "anthropic":
+            # Try live model list first, fall back to static
+            from hermes_cli.models import provider_model_ids
+            live_models = provider_model_ids("anthropic")
+            anthropic_models = live_models if live_models else [
+                "claude-opus-4-6",
+                "claude-sonnet-4-6",
+                "claude-haiku-4-5-20251001",
+            ]
+            model_choices = list(anthropic_models)
+            model_choices.append("Custom model")
+            model_choices.append(f"Keep current ({current_model})")
+
+            keep_idx = len(model_choices) - 1
+            model_idx = prompt_choice("Select default model:", model_choices, keep_idx)
+
+            if model_idx < len(anthropic_models):
+                _set_default_model(config, anthropic_models[model_idx])
+            elif model_idx == len(anthropic_models):
+                custom = prompt("Enter model name (e.g., claude-sonnet-4-20250514)")
+                if custom:
+                    _set_default_model(config, custom)
+            # else: keep current
         else:
             # Static list for OpenRouter / fallback (from canonical list)
             from hermes_cli.models import model_ids, menu_labels

pyproject.toml

@@ -13,6 +13,7 @@ license = { text = "MIT" }
 dependencies = [
   # Core
   "openai",
+  "anthropic>=0.39.0",
   "python-dotenv",
   "fire",
   "httpx",

run_agent.py

@@ -296,13 +296,16 @@ class AIAgent:
         self.base_url = base_url or OPENROUTER_BASE_URL
         provider_name = provider.strip().lower() if isinstance(provider, str) and provider.strip() else None
         self.provider = provider_name or "openrouter"
-        if api_mode in {"chat_completions", "codex_responses"}:
+        if api_mode in {"chat_completions", "codex_responses", "anthropic_messages"}:
             self.api_mode = api_mode
         elif self.provider == "openai-codex":
             self.api_mode = "codex_responses"
         elif (provider_name is None) and "chatgpt.com/backend-api/codex" in self.base_url.lower():
             self.api_mode = "codex_responses"
             self.provider = "openai-codex"
+        elif self.provider == "anthropic" or (provider_name is None and "api.anthropic.com" in self.base_url.lower()):
+            self.api_mode = "anthropic_messages"
+            self.provider = "anthropic"
         else:
             self.api_mode = "chat_completions"
 
@@ -343,7 +346,8 @@ class AIAgent:
         # conversation prefix. Uses system_and_3 strategy (4 breakpoints).
         is_openrouter = "openrouter" in self.base_url.lower()
         is_claude = "claude" in self.model.lower()
-        self._use_prompt_caching = is_openrouter and is_claude
+        is_native_anthropic = self.api_mode == "anthropic_messages"
+        self._use_prompt_caching = (is_openrouter and is_claude) or is_native_anthropic
         self._cache_ttl = "5m"  # Default 5-minute TTL (1.25x write cost)
         
         # Iteration budget pressure: warn the LLM as it approaches max_iterations.
@@ -420,66 +424,84 @@ class AIAgent:
                 ]:
                     logging.getLogger(quiet_logger).setLevel(logging.ERROR)
         
-        # Initialize OpenAI client via centralized provider router.
+        # Initialize LLM client via centralized provider router.
         # The router handles auth resolution, base URL, headers, and
-        # Codex wrapping for all known providers.
+        # Codex/Anthropic wrapping for all known providers.
         # raw_codex=True because the main agent needs direct responses.stream()
         # access for Codex Responses API streaming.
-        if api_key and base_url:
-            # Explicit credentials from CLI/gateway — construct directly.
-            # The runtime provider resolver already handled auth for us.
-            client_kwargs = {"api_key": api_key, "base_url": base_url}
-            effective_base = base_url
-            if "openrouter" in effective_base.lower():
-                client_kwargs["default_headers"] = {
-                    "HTTP-Referer": "https://hermes-agent.nousresearch.com",
-                    "X-OpenRouter-Title": "Hermes Agent",
-                    "X-OpenRouter-Categories": "productivity,cli-agent",
-                }
-            elif "api.kimi.com" in effective_base.lower():
-                client_kwargs["default_headers"] = {
-                    "User-Agent": "KimiCLI/1.3",
-                }
+        self._anthropic_client = None
+
+        if self.api_mode == "anthropic_messages":
+            from agent.anthropic_adapter import build_anthropic_client
+            effective_key = api_key or os.getenv("ANTHROPIC_API_KEY", "") or os.getenv("ANTHROPIC_TOKEN", "")
+            if not effective_key:
+                from agent.anthropic_adapter import resolve_anthropic_token
+                effective_key = resolve_anthropic_token() or ""
+            self._anthropic_api_key = effective_key
+            self._anthropic_client = build_anthropic_client(effective_key, base_url if base_url and "anthropic" in base_url else None)
+            # No OpenAI client needed for Anthropic mode
+            self.client = None
+            self._client_kwargs = {}
+            if not self.quiet_mode:
+                print(f"🤖 AI Agent initialized with model: {self.model} (Anthropic native)")
+                if effective_key and len(effective_key) > 12:
+                    print(f"🔑 Using token: {effective_key[:8]}...{effective_key[-4:]}")
         else:
-            # No explicit creds — use the centralized provider router
-            from agent.auxiliary_client import resolve_provider_client
-            _routed_client, _ = resolve_provider_client(
-                self.provider or "auto", model=self.model, raw_codex=True)
-            if _routed_client is not None:
-                client_kwargs = {
-                    "api_key": _routed_client.api_key,
-                    "base_url": str(_routed_client.base_url),
-                }
-                # Preserve any default_headers the router set
-                if hasattr(_routed_client, '_default_headers') and _routed_client._default_headers:
-                    client_kwargs["default_headers"] = dict(_routed_client._default_headers)
-            else:
-                # Final fallback: try raw OpenRouter key
-                client_kwargs = {
-                    "api_key": os.getenv("OPENROUTER_API_KEY", ""),
-                    "base_url": OPENROUTER_BASE_URL,
-                    "default_headers": {
+            if api_key and base_url:
+                # Explicit credentials from CLI/gateway — construct directly.
+                # The runtime provider resolver already handled auth for us.
+                client_kwargs = {"api_key": api_key, "base_url": base_url}
+                effective_base = base_url
+                if "openrouter" in effective_base.lower():
+                    client_kwargs["default_headers"] = {
                         "HTTP-Referer": "https://hermes-agent.nousresearch.com",
                         "X-OpenRouter-Title": "Hermes Agent",
                         "X-OpenRouter-Categories": "productivity,cli-agent",
-                    },
-                }
-        
-        self._client_kwargs = client_kwargs  # stored for rebuilding after interrupt
-        try:
-            self.client = OpenAI(**client_kwargs)
-            if not self.quiet_mode:
-                print(f"🤖 AI Agent initialized with model: {self.model}")
-                if base_url:
-                    print(f"🔗 Using custom base URL: {base_url}")
-                # Always show API key info (masked) for debugging auth issues
-                key_used = client_kwargs.get("api_key", "none")
-                if key_used and key_used != "dummy-key" and len(key_used) > 12:
-                    print(f"🔑 Using API key: {key_used[:8]}...{key_used[-4:]}")
+                    }
+                elif "api.kimi.com" in effective_base.lower():
+                    client_kwargs["default_headers"] = {
+                        "User-Agent": "KimiCLI/1.3",
+                    }
+            else:
+                # No explicit creds — use the centralized provider router
+                from agent.auxiliary_client import resolve_provider_client
+                _routed_client, _ = resolve_provider_client(
+                    self.provider or "auto", model=self.model, raw_codex=True)
+                if _routed_client is not None:
+                    client_kwargs = {
+                        "api_key": _routed_client.api_key,
+                        "base_url": str(_routed_client.base_url),
+                    }
+                    # Preserve any default_headers the router set
+                    if hasattr(_routed_client, '_default_headers') and _routed_client._default_headers:
+                        client_kwargs["default_headers"] = dict(_routed_client._default_headers)
                 else:
-                    print(f"⚠️  Warning: API key appears invalid or missing (got: '{key_used[:20] if key_used else 'none'}...')")
-        except Exception as e:
-            raise RuntimeError(f"Failed to initialize OpenAI client: {e}")
+                    # Final fallback: try raw OpenRouter key
+                    client_kwargs = {
+                        "api_key": os.getenv("OPENROUTER_API_KEY", ""),
+                        "base_url": OPENROUTER_BASE_URL,
+                        "default_headers": {
+                            "HTTP-Referer": "https://hermes-agent.nousresearch.com",
+                            "X-OpenRouter-Title": "Hermes Agent",
+                            "X-OpenRouter-Categories": "productivity,cli-agent",
+                        },
+                    }
+            
+            self._client_kwargs = client_kwargs  # stored for rebuilding after interrupt
+            try:
+                self.client = OpenAI(**client_kwargs)
+                if not self.quiet_mode:
+                    print(f"🤖 AI Agent initialized with model: {self.model}")
+                    if base_url:
+                        print(f"🔗 Using custom base URL: {base_url}")
+                    # Always show API key info (masked) for debugging auth issues
+                    key_used = client_kwargs.get("api_key", "none")
+                    if key_used and key_used != "dummy-key" and len(key_used) > 12:
+                        print(f"🔑 Using API key: {key_used[:8]}...{key_used[-4:]}")
+                    else:
+                        print(f"⚠️  Warning: API key appears invalid or missing (got: '{key_used[:20] if key_used else 'none'}...')")
+            except Exception as e:
+                raise RuntimeError(f"Failed to initialize OpenAI client: {e}")
         
         # Provider fallback — a single backup model/provider tried when the
         # primary is exhausted (rate-limit, overload, connection failure).
@@ -533,7 +555,8 @@ class AIAgent:
         
         # Show prompt caching status
         if self._use_prompt_caching and not self.quiet_mode:
-            print(f"💾 Prompt caching: ENABLED (Claude via OpenRouter, {self._cache_ttl} TTL)")
+            source = "native Anthropic" if is_native_anthropic else "Claude via OpenRouter"
+            print(f"💾 Prompt caching: ENABLED ({source}, {self._cache_ttl} TTL)")
         
         # Session logging setup - auto-save conversation trajectories for debugging
         self.session_start = datetime.now()
@@ -2233,6 +2256,8 @@ class AIAgent:
             try:
                 if self.api_mode == "codex_responses":
                     result["response"] = self._run_codex_stream(api_kwargs)
+                elif self.api_mode == "anthropic_messages":
+                    result["response"] = self._anthropic_client.messages.create(**api_kwargs)
                 else:
                     result["response"] = self.client.chat.completions.create(**api_kwargs)
             except Exception as e:
@@ -2245,12 +2270,19 @@ class AIAgent:
             if self._interrupt_requested:
                 # Force-close the HTTP connection to stop token generation
                 try:
-                    self.client.close()
+                    if self.api_mode == "anthropic_messages":
+                        self._anthropic_client.close()
+                    else:
+                        self.client.close()
                 except Exception:
                     pass
                 # Rebuild the client for future calls (cheap, no network)
                 try:
-                    self.client = OpenAI(**self._client_kwargs)
+                    if self.api_mode == "anthropic_messages":
+                        from agent.anthropic_adapter import build_anthropic_client
+                        self._anthropic_client = build_anthropic_client(self._anthropic_api_key)
+                    else:
+                        self.client = OpenAI(**self._client_kwargs)
                 except Exception:
                     pass
                 raise InterruptedError("Agent interrupted during API call")
@@ -2336,6 +2368,16 @@ class AIAgent:
 
     def _build_api_kwargs(self, api_messages: list) -> dict:
         """Build the keyword arguments dict for the active API mode."""
+        if self.api_mode == "anthropic_messages":
+            from agent.anthropic_adapter import build_anthropic_kwargs
+            return build_anthropic_kwargs(
+                model=self.model,
+                messages=api_messages,
+                tools=self.tools,
+                max_tokens=None,
+                reasoning_config=self.reasoning_config,
+            )
+
         if self.api_mode == "codex_responses":
             instructions = ""
             payload_messages = api_messages
@@ -2659,6 +2701,15 @@ class AIAgent:
                 if "max_output_tokens" in codex_kwargs:
                     codex_kwargs["max_output_tokens"] = 5120
                 response = self._run_codex_stream(codex_kwargs)
+            elif not _aux_available and self.api_mode == "anthropic_messages":
+                # Native Anthropic — use the Anthropic client directly
+                from agent.anthropic_adapter import build_anthropic_kwargs as _build_ant_kwargs
+                ant_kwargs = _build_ant_kwargs(
+                    model=self.model, messages=api_messages,
+                    tools=[memory_tool_def], max_tokens=5120,
+                    reasoning_config=None,
+                )
+                response = self._anthropic_client.messages.create(**ant_kwargs)
             elif not _aux_available:
                 api_kwargs = {
                     "model": self.model,
@@ -2669,12 +2720,17 @@ class AIAgent:
                 }
                 response = self.client.chat.completions.create(**api_kwargs, timeout=30.0)
 
-            # Extract tool calls from the response, handling both API formats
+            # Extract tool calls from the response, handling all API formats
             tool_calls = []
             if self.api_mode == "codex_responses" and not _aux_available:
                 assistant_msg, _ = self._normalize_codex_response(response)
                 if assistant_msg and assistant_msg.tool_calls:
                     tool_calls = assistant_msg.tool_calls
+            elif self.api_mode == "anthropic_messages" and not _aux_available:
+                from agent.anthropic_adapter import normalize_anthropic_response as _nar_flush
+                _flush_msg, _ = _nar_flush(response)
+                if _flush_msg and _flush_msg.tool_calls:
+                    tool_calls = _flush_msg.tool_calls
             elif hasattr(response, "choices") and response.choices:
                 assistant_message = response.choices[0].message
                 if assistant_message.tool_calls:
@@ -3147,12 +3203,20 @@ class AIAgent:
                 if summary_extra_body:
                     summary_kwargs["extra_body"] = summary_extra_body
 
-                summary_response = self.client.chat.completions.create(**summary_kwargs)
-
-                if summary_response.choices and summary_response.choices[0].message.content:
-                    final_response = summary_response.choices[0].message.content
+                if self.api_mode == "anthropic_messages":
+                    from agent.anthropic_adapter import build_anthropic_kwargs as _bak, normalize_anthropic_response as _nar
+                    _ant_kw = _bak(model=self.model, messages=api_messages, tools=None,
+                                   max_tokens=self.max_tokens, reasoning_config=self.reasoning_config)
+                    summary_response = self._anthropic_client.messages.create(**_ant_kw)
+                    _msg, _ = _nar(summary_response)
+                    final_response = (_msg.content or "").strip()
                 else:
-                    final_response = ""
+                    summary_response = self.client.chat.completions.create(**summary_kwargs)
+
+                    if summary_response.choices and summary_response.choices[0].message.content:
+                        final_response = summary_response.choices[0].message.content
+                    else:
+                        final_response = ""
 
             if final_response:
                 if "<think>" in final_response:
@@ -3169,6 +3233,13 @@ class AIAgent:
                     retry_response = self._run_codex_stream(codex_kwargs)
                     retry_msg, _ = self._normalize_codex_response(retry_response)
                     final_response = (retry_msg.content or "").strip() if retry_msg else ""
+                elif self.api_mode == "anthropic_messages":
+                    from agent.anthropic_adapter import build_anthropic_kwargs as _bak2, normalize_anthropic_response as _nar2
+                    _ant_kw2 = _bak2(model=self.model, messages=api_messages, tools=None,
+                                     max_tokens=self.max_tokens, reasoning_config=self.reasoning_config)
+                    retry_response = self._anthropic_client.messages.create(**_ant_kw2)
+                    _retry_msg, _ = _nar2(retry_response)
+                    final_response = (_retry_msg.content or "").strip()
                 else:
                     summary_kwargs = {
                         "model": self.model,
@@ -3548,6 +3619,7 @@ class AIAgent:
             compression_attempts = 0
             max_compression_attempts = 3
             codex_auth_retry_attempted = False
+            anthropic_auth_retry_attempted = False
             nous_auth_retry_attempted = False
             restart_with_compressed_messages = False
             restart_with_length_continuation = False
@@ -3598,6 +3670,17 @@ class AIAgent:
                         elif len(output_items) == 0:
                             response_invalid = True
                             error_details.append("response.output is empty")
+                    elif self.api_mode == "anthropic_messages":
+                        content_blocks = getattr(response, "content", None) if response is not None else None
+                        if response is None:
+                            response_invalid = True
+                            error_details.append("response is None")
+                        elif not isinstance(content_blocks, list):
+                            response_invalid = True
+                            error_details.append("response.content is not a list")
+                        elif len(content_blocks) == 0:
+                            response_invalid = True
+                            error_details.append("response.content is empty")
                     else:
                         if response is None or not hasattr(response, 'choices') or response.choices is None or len(response.choices) == 0:
                             response_invalid = True
@@ -3699,6 +3782,9 @@ class AIAgent:
                             finish_reason = "length"
                         else:
                             finish_reason = "stop"
+                    elif self.api_mode == "anthropic_messages":
+                        stop_reason_map = {"end_turn": "stop", "tool_use": "tool_calls", "max_tokens": "length", "stop_sequence": "stop"}
+                        finish_reason = stop_reason_map.get(response.stop_reason, "stop")
                     else:
                         finish_reason = response.choices[0].finish_reason
 
@@ -3776,7 +3862,7 @@ class AIAgent:
                     
                     # Track actual token usage from response for context management
                     if hasattr(response, 'usage') and response.usage:
-                        if self.api_mode == "codex_responses":
+                        if self.api_mode in ("codex_responses", "anthropic_messages"):
                             prompt_tokens = getattr(response.usage, 'input_tokens', 0) or 0
                             completion_tokens = getattr(response.usage, 'output_tokens', 0) or 0
                             total_tokens = (
@@ -3811,9 +3897,15 @@ class AIAgent:
                         
                         # Log cache hit stats when prompt caching is active
                         if self._use_prompt_caching:
-                            details = getattr(response.usage, 'prompt_tokens_details', None)
-                            cached = getattr(details, 'cached_tokens', 0) or 0 if details else 0
-                            written = getattr(details, 'cache_write_tokens', 0) or 0 if details else 0
+                            if self.api_mode == "anthropic_messages":
+                                # Anthropic uses cache_read_input_tokens / cache_creation_input_tokens
+                                cached = getattr(response.usage, 'cache_read_input_tokens', 0) or 0
+                                written = getattr(response.usage, 'cache_creation_input_tokens', 0) or 0
+                            else:
+                                # OpenRouter uses prompt_tokens_details.cached_tokens
+                                details = getattr(response.usage, 'prompt_tokens_details', None)
+                                cached = getattr(details, 'cached_tokens', 0) or 0 if details else 0
+                                written = getattr(details, 'cache_write_tokens', 0) or 0 if details else 0
                             prompt = usage_dict["prompt_tokens"]
                             hit_pct = (cached / prompt * 100) if prompt > 0 else 0
                             if not self.quiet_mode:
@@ -3863,6 +3955,21 @@ class AIAgent:
                         if self._try_refresh_nous_client_credentials(force=True):
                             print(f"{self.log_prefix}🔐 Nous agent key refreshed after 401. Retrying request...")
                             continue
+                    if (
+                        self.api_mode == "anthropic_messages"
+                        and status_code == 401
+                        and hasattr(self, '_anthropic_api_key')
+                        and not anthropic_auth_retry_attempted
+                    ):
+                        anthropic_auth_retry_attempted = True
+                        # Try re-reading Claude Code credentials (they may have been refreshed)
+                        from agent.anthropic_adapter import resolve_anthropic_token, build_anthropic_client
+                        new_token = resolve_anthropic_token()
+                        if new_token and new_token != self._anthropic_api_key:
+                            self._anthropic_api_key = new_token
+                            self._anthropic_client = build_anthropic_client(new_token)
+                            print(f"{self.log_prefix}🔐 Anthropic credentials refreshed after 401. Retrying request...")
+                            continue
 
                     retry_count += 1
                     elapsed_time = time.time() - api_start_time
@@ -4105,6 +4212,9 @@ class AIAgent:
             try:
                 if self.api_mode == "codex_responses":
                     assistant_message, finish_reason = self._normalize_codex_response(response)
+                elif self.api_mode == "anthropic_messages":
+                    from agent.anthropic_adapter import normalize_anthropic_response
+                    assistant_message, finish_reason = normalize_anthropic_response(response)
                 else:
                     assistant_message = response.choices[0].message
                 

tests/test_anthropic_adapter.py

@@ -0,0 +1,500 @@
+"""Tests for agent/anthropic_adapter.py — Anthropic Messages API adapter."""
+
+import json
+import time
+from types import SimpleNamespace
+from unittest.mock import patch, MagicMock
+
+import pytest
+
+from agent.anthropic_adapter import (
+    _is_oauth_token,
+    build_anthropic_client,
+    build_anthropic_kwargs,
+    convert_messages_to_anthropic,
+    convert_tools_to_anthropic,
+    is_claude_code_token_valid,
+    normalize_anthropic_response,
+    normalize_model_name,
+    read_claude_code_credentials,
+    resolve_anthropic_token,
+)
+
+
+# ---------------------------------------------------------------------------
+# Auth helpers
+# ---------------------------------------------------------------------------
+
+
+class TestIsOAuthToken:
+    def test_setup_token(self):
+        assert _is_oauth_token("sk-ant-oat01-abcdef1234567890") is True
+
+    def test_api_key(self):
+        assert _is_oauth_token("sk-ant-api03-abcdef1234567890") is False
+
+    def test_managed_key(self):
+        # Managed keys from ~/.claude.json are NOT regular API keys
+        assert _is_oauth_token("ou1R1z-ft0A-bDeZ9wAA") is True
+
+    def test_jwt_token(self):
+        # JWTs from OAuth flow
+        assert _is_oauth_token("eyJhbGciOiJSUzI1NiJ9.test") is True
+
+    def test_empty(self):
+        assert _is_oauth_token("") is False
+
+
+class TestBuildAnthropicClient:
+    def test_setup_token_uses_auth_token(self):
+        with patch("agent.anthropic_adapter._anthropic_sdk") as mock_sdk:
+            build_anthropic_client("sk-ant-oat01-" + "x" * 60)
+            kwargs = mock_sdk.Anthropic.call_args[1]
+            assert "auth_token" in kwargs
+            betas = kwargs["default_headers"]["anthropic-beta"]
+            assert "oauth-2025-04-20" in betas
+            assert "interleaved-thinking-2025-05-14" in betas
+            assert "fine-grained-tool-streaming-2025-05-14" in betas
+            assert "api_key" not in kwargs
+
+    def test_api_key_uses_api_key(self):
+        with patch("agent.anthropic_adapter._anthropic_sdk") as mock_sdk:
+            build_anthropic_client("sk-ant-api03-something")
+            kwargs = mock_sdk.Anthropic.call_args[1]
+            assert kwargs["api_key"] == "sk-ant-api03-something"
+            assert "auth_token" not in kwargs
+            # API key auth should still get common betas
+            betas = kwargs["default_headers"]["anthropic-beta"]
+            assert "interleaved-thinking-2025-05-14" in betas
+            assert "oauth-2025-04-20" not in betas  # OAuth-only beta NOT present
+
+    def test_custom_base_url(self):
+        with patch("agent.anthropic_adapter._anthropic_sdk") as mock_sdk:
+            build_anthropic_client("sk-ant-api03-x", base_url="https://custom.api.com")
+            kwargs = mock_sdk.Anthropic.call_args[1]
+            assert kwargs["base_url"] == "https://custom.api.com"
+
+
+class TestReadClaudeCodeCredentials:
+    def test_reads_valid_credentials(self, tmp_path, monkeypatch):
+        cred_file = tmp_path / ".claude" / ".credentials.json"
+        cred_file.parent.mkdir(parents=True)
+        cred_file.write_text(json.dumps({
+            "claudeAiOauth": {
+                "accessToken": "sk-ant-oat01-test-token",
+                "refreshToken": "sk-ant-ort01-refresh",
+                "expiresAt": int(time.time() * 1000) + 3600_000,
+            }
+        }))
+        monkeypatch.setattr("agent.anthropic_adapter.Path.home", lambda: tmp_path)
+        creds = read_claude_code_credentials()
+        assert creds is not None
+        assert creds["accessToken"] == "sk-ant-oat01-test-token"
+        assert creds["refreshToken"] == "sk-ant-ort01-refresh"
+
+    def test_returns_none_for_missing_file(self, tmp_path, monkeypatch):
+        monkeypatch.setattr("agent.anthropic_adapter.Path.home", lambda: tmp_path)
+        assert read_claude_code_credentials() is None
+
+    def test_returns_none_for_missing_oauth_key(self, tmp_path, monkeypatch):
+        cred_file = tmp_path / ".claude" / ".credentials.json"
+        cred_file.parent.mkdir(parents=True)
+        cred_file.write_text(json.dumps({"someOtherKey": {}}))
+        monkeypatch.setattr("agent.anthropic_adapter.Path.home", lambda: tmp_path)
+        assert read_claude_code_credentials() is None
+
+    def test_returns_none_for_empty_access_token(self, tmp_path, monkeypatch):
+        cred_file = tmp_path / ".claude" / ".credentials.json"
+        cred_file.parent.mkdir(parents=True)
+        cred_file.write_text(json.dumps({
+            "claudeAiOauth": {"accessToken": "", "refreshToken": "x"}
+        }))
+        monkeypatch.setattr("agent.anthropic_adapter.Path.home", lambda: tmp_path)
+        assert read_claude_code_credentials() is None
+
+
+class TestIsClaudeCodeTokenValid:
+    def test_valid_token(self):
+        creds = {"accessToken": "tok", "expiresAt": int(time.time() * 1000) + 3600_000}
+        assert is_claude_code_token_valid(creds) is True
+
+    def test_expired_token(self):
+        creds = {"accessToken": "tok", "expiresAt": int(time.time() * 1000) - 3600_000}
+        assert is_claude_code_token_valid(creds) is False
+
+    def test_no_expiry_but_has_token(self):
+        creds = {"accessToken": "tok", "expiresAt": 0}
+        assert is_claude_code_token_valid(creds) is True
+
+
+class TestResolveAnthropicToken:
+    def test_prefers_api_key(self, monkeypatch):
+        monkeypatch.setenv("ANTHROPIC_API_KEY", "sk-ant-api03-mykey")
+        monkeypatch.setenv("ANTHROPIC_TOKEN", "sk-ant-oat01-mytoken")
+        assert resolve_anthropic_token() == "sk-ant-api03-mykey"
+
+    def test_falls_back_to_token(self, monkeypatch):
+        monkeypatch.delenv("ANTHROPIC_API_KEY", raising=False)
+        monkeypatch.setenv("ANTHROPIC_TOKEN", "sk-ant-oat01-mytoken")
+        assert resolve_anthropic_token() == "sk-ant-oat01-mytoken"
+
+    def test_returns_none_with_no_creds(self, monkeypatch, tmp_path):
+        monkeypatch.delenv("ANTHROPIC_API_KEY", raising=False)
+        monkeypatch.delenv("ANTHROPIC_TOKEN", raising=False)
+        monkeypatch.delenv("CLAUDE_CODE_OAUTH_TOKEN", raising=False)
+        monkeypatch.setattr("agent.anthropic_adapter.Path.home", lambda: tmp_path)
+        assert resolve_anthropic_token() is None
+
+
+# ---------------------------------------------------------------------------
+# Model name normalization
+# ---------------------------------------------------------------------------
+
+
+class TestNormalizeModelName:
+    def test_strips_anthropic_prefix(self):
+        assert normalize_model_name("anthropic/claude-sonnet-4-20250514") == "claude-sonnet-4-20250514"
+
+    def test_leaves_bare_name(self):
+        assert normalize_model_name("claude-sonnet-4-20250514") == "claude-sonnet-4-20250514"
+
+
+# ---------------------------------------------------------------------------
+# Tool conversion
+# ---------------------------------------------------------------------------
+
+
+class TestConvertTools:
+    def test_converts_openai_to_anthropic_format(self):
+        tools = [
+            {
+                "type": "function",
+                "function": {
+                    "name": "search",
+                    "description": "Search the web",
+                    "parameters": {
+                        "type": "object",
+                        "properties": {"query": {"type": "string"}},
+                        "required": ["query"],
+                    },
+                },
+            }
+        ]
+        result = convert_tools_to_anthropic(tools)
+        assert len(result) == 1
+        assert result[0]["name"] == "search"
+        assert result[0]["description"] == "Search the web"
+        assert result[0]["input_schema"]["properties"]["query"]["type"] == "string"
+
+    def test_empty_tools(self):
+        assert convert_tools_to_anthropic([]) == []
+        assert convert_tools_to_anthropic(None) == []
+
+
+# ---------------------------------------------------------------------------
+# Message conversion
+# ---------------------------------------------------------------------------
+
+
+class TestConvertMessages:
+    def test_extracts_system_prompt(self):
+        messages = [
+            {"role": "system", "content": "You are helpful."},
+            {"role": "user", "content": "Hello"},
+        ]
+        system, result = convert_messages_to_anthropic(messages)
+        assert system == "You are helpful."
+        assert len(result) == 1
+        assert result[0]["role"] == "user"
+
+    def test_converts_tool_calls(self):
+        messages = [
+            {
+                "role": "assistant",
+                "content": "Let me search.",
+                "tool_calls": [
+                    {
+                        "id": "tc_1",
+                        "function": {
+                            "name": "search",
+                            "arguments": '{"query": "test"}',
+                        },
+                    }
+                ],
+            },
+            {"role": "tool", "tool_call_id": "tc_1", "content": "search results"},
+        ]
+        _, result = convert_messages_to_anthropic(messages)
+        blocks = result[0]["content"]
+        assert blocks[0] == {"type": "text", "text": "Let me search."}
+        assert blocks[1]["type"] == "tool_use"
+        assert blocks[1]["id"] == "tc_1"
+        assert blocks[1]["input"] == {"query": "test"}
+
+    def test_converts_tool_results(self):
+        messages = [
+            {"role": "tool", "tool_call_id": "tc_1", "content": "result data"},
+        ]
+        _, result = convert_messages_to_anthropic(messages)
+        assert result[0]["role"] == "user"
+        assert result[0]["content"][0]["type"] == "tool_result"
+        assert result[0]["content"][0]["tool_use_id"] == "tc_1"
+
+    def test_merges_consecutive_tool_results(self):
+        messages = [
+            {"role": "tool", "tool_call_id": "tc_1", "content": "result 1"},
+            {"role": "tool", "tool_call_id": "tc_2", "content": "result 2"},
+        ]
+        _, result = convert_messages_to_anthropic(messages)
+        assert len(result) == 1
+        assert len(result[0]["content"]) == 2
+
+    def test_strips_orphaned_tool_use(self):
+        messages = [
+            {
+                "role": "assistant",
+                "content": "",
+                "tool_calls": [
+                    {"id": "tc_orphan", "function": {"name": "x", "arguments": "{}"}}
+                ],
+            },
+            {"role": "user", "content": "never mind"},
+        ]
+        _, result = convert_messages_to_anthropic(messages)
+        # tc_orphan has no matching tool_result, should be stripped
+        assistant_blocks = result[0]["content"]
+        assert all(b.get("type") != "tool_use" for b in assistant_blocks)
+
+    def test_system_with_cache_control(self):
+        messages = [
+            {
+                "role": "system",
+                "content": [
+                    {"type": "text", "text": "System prompt", "cache_control": {"type": "ephemeral"}},
+                ],
+            },
+            {"role": "user", "content": "Hi"},
+        ]
+        system, result = convert_messages_to_anthropic(messages)
+        # When cache_control is present, system should be a list of blocks
+        assert isinstance(system, list)
+        assert system[0]["cache_control"] == {"type": "ephemeral"}
+
+
+# ---------------------------------------------------------------------------
+# Build kwargs
+# ---------------------------------------------------------------------------
+
+
+class TestBuildAnthropicKwargs:
+    def test_basic_kwargs(self):
+        messages = [
+            {"role": "system", "content": "Be helpful."},
+            {"role": "user", "content": "Hi"},
+        ]
+        kwargs = build_anthropic_kwargs(
+            model="claude-sonnet-4-20250514",
+            messages=messages,
+            tools=None,
+            max_tokens=4096,
+            reasoning_config=None,
+        )
+        assert kwargs["model"] == "claude-sonnet-4-20250514"
+        assert kwargs["system"] == "Be helpful."
+        assert kwargs["max_tokens"] == 4096
+        assert "tools" not in kwargs
+
+    def test_strips_anthropic_prefix(self):
+        kwargs = build_anthropic_kwargs(
+            model="anthropic/claude-sonnet-4-20250514",
+            messages=[{"role": "user", "content": "Hi"}],
+            tools=None,
+            max_tokens=4096,
+            reasoning_config=None,
+        )
+        assert kwargs["model"] == "claude-sonnet-4-20250514"
+
+    def test_reasoning_config_maps_to_thinking(self):
+        kwargs = build_anthropic_kwargs(
+            model="claude-sonnet-4-20250514",
+            messages=[{"role": "user", "content": "think hard"}],
+            tools=None,
+            max_tokens=4096,
+            reasoning_config={"enabled": True, "effort": "high"},
+        )
+        assert kwargs["thinking"]["type"] == "enabled"
+        assert kwargs["thinking"]["budget_tokens"] == 16000
+        assert kwargs["max_tokens"] >= 16000 + 4096
+
+    def test_reasoning_disabled(self):
+        kwargs = build_anthropic_kwargs(
+            model="claude-sonnet-4-20250514",
+            messages=[{"role": "user", "content": "quick"}],
+            tools=None,
+            max_tokens=4096,
+            reasoning_config={"enabled": False},
+        )
+        assert "thinking" not in kwargs
+
+    def test_default_max_tokens(self):
+        kwargs = build_anthropic_kwargs(
+            model="claude-sonnet-4-20250514",
+            messages=[{"role": "user", "content": "Hi"}],
+            tools=None,
+            max_tokens=None,
+            reasoning_config=None,
+        )
+        assert kwargs["max_tokens"] == 16384
+
+
+# ---------------------------------------------------------------------------
+# Response normalization
+# ---------------------------------------------------------------------------
+
+
+class TestNormalizeResponse:
+    def _make_response(self, content_blocks, stop_reason="end_turn"):
+        resp = SimpleNamespace()
+        resp.content = content_blocks
+        resp.stop_reason = stop_reason
+        resp.usage = SimpleNamespace(input_tokens=100, output_tokens=50)
+        return resp
+
+    def test_text_response(self):
+        block = SimpleNamespace(type="text", text="Hello world")
+        msg, reason = normalize_anthropic_response(self._make_response([block]))
+        assert msg.content == "Hello world"
+        assert reason == "stop"
+        assert msg.tool_calls is None
+
+    def test_tool_use_response(self):
+        blocks = [
+            SimpleNamespace(type="text", text="Searching..."),
+            SimpleNamespace(
+                type="tool_use",
+                id="tc_1",
+                name="search",
+                input={"query": "test"},
+            ),
+        ]
+        msg, reason = normalize_anthropic_response(
+            self._make_response(blocks, "tool_use")
+        )
+        assert msg.content == "Searching..."
+        assert reason == "tool_calls"
+        assert len(msg.tool_calls) == 1
+        assert msg.tool_calls[0].function.name == "search"
+        assert json.loads(msg.tool_calls[0].function.arguments) == {"query": "test"}
+
+    def test_thinking_response(self):
+        blocks = [
+            SimpleNamespace(type="thinking", thinking="Let me reason about this..."),
+            SimpleNamespace(type="text", text="The answer is 42."),
+        ]
+        msg, reason = normalize_anthropic_response(self._make_response(blocks))
+        assert msg.content == "The answer is 42."
+        assert msg.reasoning == "Let me reason about this..."
+
+    def test_stop_reason_mapping(self):
+        block = SimpleNamespace(type="text", text="x")
+        _, r1 = normalize_anthropic_response(
+            self._make_response([block], "end_turn")
+        )
+        _, r2 = normalize_anthropic_response(
+            self._make_response([block], "tool_use")
+        )
+        _, r3 = normalize_anthropic_response(
+            self._make_response([block], "max_tokens")
+        )
+        assert r1 == "stop"
+        assert r2 == "tool_calls"
+        assert r3 == "length"
+
+    def test_no_text_content(self):
+        block = SimpleNamespace(
+            type="tool_use", id="tc_1", name="search", input={"q": "hi"}
+        )
+        msg, reason = normalize_anthropic_response(
+            self._make_response([block], "tool_use")
+        )
+        assert msg.content is None
+        assert len(msg.tool_calls) == 1
+
+
+# ---------------------------------------------------------------------------
+# Role alternation
+# ---------------------------------------------------------------------------
+
+
+class TestRoleAlternation:
+    def test_merges_consecutive_user_messages(self):
+        messages = [
+            {"role": "user", "content": "Hello"},
+            {"role": "user", "content": "World"},
+        ]
+        _, result = convert_messages_to_anthropic(messages)
+        assert len(result) == 1
+        assert result[0]["role"] == "user"
+        assert "Hello" in result[0]["content"]
+        assert "World" in result[0]["content"]
+
+    def test_preserves_proper_alternation(self):
+        messages = [
+            {"role": "user", "content": "Hi"},
+            {"role": "assistant", "content": "Hello!"},
+            {"role": "user", "content": "How are you?"},
+        ]
+        _, result = convert_messages_to_anthropic(messages)
+        assert len(result) == 3
+        assert [m["role"] for m in result] == ["user", "assistant", "user"]
+
+
+# ---------------------------------------------------------------------------
+# Tool choice
+# ---------------------------------------------------------------------------
+
+
+class TestToolChoice:
+    _DUMMY_TOOL = [
+        {
+            "type": "function",
+            "function": {
+                "name": "test",
+                "description": "x",
+                "parameters": {"type": "object", "properties": {}},
+            },
+        }
+    ]
+
+    def test_auto_tool_choice(self):
+        kwargs = build_anthropic_kwargs(
+            model="claude-sonnet-4-20250514",
+            messages=[{"role": "user", "content": "Hi"}],
+            tools=self._DUMMY_TOOL,
+            max_tokens=4096,
+            reasoning_config=None,
+            tool_choice="auto",
+        )
+        assert kwargs["tool_choice"] == {"type": "auto"}
+
+    def test_required_tool_choice(self):
+        kwargs = build_anthropic_kwargs(
+            model="claude-sonnet-4-20250514",
+            messages=[{"role": "user", "content": "Hi"}],
+            tools=self._DUMMY_TOOL,
+            max_tokens=4096,
+            reasoning_config=None,
+            tool_choice="required",
+        )
+        assert kwargs["tool_choice"] == {"type": "any"}
+
+    def test_specific_tool_choice(self):
+        kwargs = build_anthropic_kwargs(
+            model="claude-sonnet-4-20250514",
+            messages=[{"role": "user", "content": "Hi"}],
+            tools=self._DUMMY_TOOL,
+            max_tokens=4096,
+            reasoning_config=None,
+            tool_choice="search",
+        )
+        assert kwargs["tool_choice"] == {"type": "tool", "name": "search"}

tests/test_run_agent.py

@@ -281,20 +281,21 @@ class TestMaskApiKey:
 
 class TestInit:
     def test_anthropic_base_url_accepted(self):
-        """Anthropic base URLs should be accepted (OpenAI-compatible endpoint)."""
+        """Anthropic base URLs should route to native Anthropic client."""
         with (
             patch("run_agent.get_tool_definitions", return_value=[]),
             patch("run_agent.check_toolset_requirements", return_value={}),
-            patch("run_agent.OpenAI") as mock_openai,
+            patch("agent.anthropic_adapter._anthropic_sdk") as mock_anthropic,
         ):
-            AIAgent(
+            agent = AIAgent(
                 api_key="test-key-1234567890",
                 base_url="https://api.anthropic.com/v1/",
                 quiet_mode=True,
                 skip_context_files=True,
                 skip_memory=True,
             )
-            mock_openai.assert_called_once()
+            assert agent.api_mode == "anthropic_messages"
+            mock_anthropic.Anthropic.assert_called_once()
 
     def test_prompt_caching_claude_openrouter(self):
         """Claude model via OpenRouter should enable prompt caching."""
@@ -345,6 +346,23 @@ class TestInit:
             )
             assert a._use_prompt_caching is False
 
+    def test_prompt_caching_native_anthropic(self):
+        """Native Anthropic provider should enable prompt caching."""
+        with (
+            patch("run_agent.get_tool_definitions", return_value=[]),
+            patch("run_agent.check_toolset_requirements", return_value={}),
+            patch("agent.anthropic_adapter._anthropic_sdk"),
+        ):
+            a = AIAgent(
+                api_key="test-key-1234567890",
+                base_url="https://api.anthropic.com/v1/",
+                quiet_mode=True,
+                skip_context_files=True,
+                skip_memory=True,
+            )
+            assert a.api_mode == "anthropic_messages"
+            assert a._use_prompt_caching is True
+
     def test_valid_tool_names_populated(self):
         """valid_tool_names should contain names from loaded tools."""
         tools = _make_tool_defs("web_search", "terminal")

website/docs/getting-started/quickstart.md

@@ -43,6 +43,7 @@ hermes setup       # Or configure everything at once
 |----------|-----------|---------------|
 | **Nous Portal** | Subscription-based, zero-config | OAuth login via `hermes model` |
 | **OpenAI Codex** | ChatGPT OAuth, uses Codex models | Device code auth via `hermes model` |
+| **Anthropic** | Claude models directly (Pro/Max or API key) | API key or Claude Code setup-token |
 | **OpenRouter** | 200+ models, pay-per-use | Enter your API key |
 | **Custom Endpoint** | VLLM, SGLang, any OpenAI-compatible API | Set base URL + API key |
 

website/docs/reference/environment-variables.md

@@ -23,6 +23,9 @@ All variables go in `~/.hermes/.env`. You can also set them with `hermes config
 | `MINIMAX_BASE_URL` | Override MiniMax base URL (default: `https://api.minimax.io/v1`) |
 | `MINIMAX_CN_API_KEY` | MiniMax API key — China endpoint ([minimaxi.com](https://www.minimaxi.com)) |
 | `MINIMAX_CN_BASE_URL` | Override MiniMax China base URL (default: `https://api.minimaxi.com/v1`) |
+| `ANTHROPIC_API_KEY` | Anthropic API key or setup-token ([console.anthropic.com](https://console.anthropic.com/)) |
+| `ANTHROPIC_TOKEN` | Anthropic OAuth/setup token (alternative to `ANTHROPIC_API_KEY`) |
+| `CLAUDE_CODE_OAUTH_TOKEN` | Claude Code setup-token (same as `ANTHROPIC_TOKEN`) |
 | `HERMES_MODEL` | Preferred model name (checked before `LLM_MODEL`, used by gateway) |
 | `LLM_MODEL` | Default model name (fallback when not set in config.yaml) |
 | `VOICE_TOOLS_OPENAI_KEY` | OpenAI key for TTS and voice transcription (separate from custom endpoint) |
@@ -32,7 +35,7 @@ All variables go in `~/.hermes/.env`. You can also set them with `hermes config
 
 | Variable | Description |
 |----------|-------------|
-| `HERMES_INFERENCE_PROVIDER` | Override provider selection: `auto`, `openrouter`, `nous`, `zai`, `kimi-coding`, `minimax`, `minimax-cn` (default: `auto`) |
+| `HERMES_INFERENCE_PROVIDER` | Override provider selection: `auto`, `openrouter`, `nous`, `anthropic`, `zai`, `kimi-coding`, `minimax`, `minimax-cn` (default: `auto`) |
 | `HERMES_PORTAL_BASE_URL` | Override Nous Portal URL (for development/testing) |
 | `NOUS_INFERENCE_BASE_URL` | Override Nous inference API URL |
 | `HERMES_NOUS_MIN_KEY_TTL_SECONDS` | Min agent key TTL before re-mint (default: 1800 = 30min) |

website/docs/user-guide/configuration.md

@@ -63,6 +63,7 @@ You need at least one way to connect to an LLM. Use `hermes model` to switch pro
 |----------|-------|
 | **Nous Portal** | `hermes model` (OAuth, subscription-based) |
 | **OpenAI Codex** | `hermes model` (ChatGPT OAuth, uses Codex models) |
+| **Anthropic** | `hermes model` (API key, setup-token, or Claude Code auto-detect) |
 | **OpenRouter** | `OPENROUTER_API_KEY` in `~/.hermes/.env` |
 | **z.ai / GLM** | `GLM_API_KEY` in `~/.hermes/.env` (provider: `zai`) |
 | **Kimi / Moonshot** | `KIMI_API_KEY` in `~/.hermes/.env` (provider: `kimi-coding`) |
@@ -78,6 +79,34 @@ The OpenAI Codex provider authenticates via device code (open a URL, enter a cod
 Even when using Nous Portal, Codex, or a custom endpoint, some tools (vision, web summarization, MoA) use a separate "auxiliary" model — by default Gemini Flash via OpenRouter. An `OPENROUTER_API_KEY` enables these tools automatically. You can also configure which model and provider these tools use — see [Auxiliary Models](#auxiliary-models) below.
 :::
 
+### Anthropic (Native)
+
+Use Claude models directly through the Anthropic API — no OpenRouter proxy needed. Supports three auth methods:
+
+```bash
+# With an API key (pay-per-token)
+export ANTHROPIC_API_KEY=sk-ant-api03-...
+hermes chat --provider anthropic --model claude-sonnet-4-6
+
+# With a Claude Code setup-token (Pro/Max subscription)
+export ANTHROPIC_API_KEY=sk-ant-oat01-...  # from 'claude setup-token'
+hermes chat --provider anthropic
+
+# Auto-detect Claude Code credentials (if you have Claude Code installed)
+hermes chat --provider anthropic  # reads ~/.claude.json automatically
+```
+
+Or set it permanently:
+```yaml
+model:
+  provider: "anthropic"
+  default: "claude-sonnet-4-6"
+```
+
+:::tip Aliases
+`--provider claude` and `--provider claude-code` also work as shorthand for `--provider anthropic`.
+:::
+
 ### First-Class Chinese AI Providers
 
 These providers have built-in support with dedicated provider IDs. Set the API key and use `--provider` to select: