mirror of
https://github.com/NousResearch/hermes-agent.git
synced 2026-04-25 00:51:20 +00:00
feat: add transport ABC + AnthropicTransport wired to all paths
Add ProviderTransport ABC (4 abstract methods: convert_messages, convert_tools, build_kwargs, normalize_response) plus optional hooks (validate_response, extract_cache_stats, map_finish_reason). Add transport registry with lazy discovery — get_transport() auto-imports transport modules on first call. Add AnthropicTransport — delegates to existing anthropic_adapter.py functions, wired to ALL Anthropic code paths in run_agent.py: - Main normalize loop (L10775) - Main build_kwargs (L6673) - Response validation (L9366) - Finish reason mapping (L9534) - Cache stats extraction (L9827) - Truncation normalize (L9565) - Memory flush build_kwargs + normalize (L7363, L7395) - Iteration-limit summary + retry (L8465, L8498) Zero direct adapter imports remain for transport methods. Client lifecycle, streaming, auth, and credential management stay on AIAgent. 20 new tests (ABC contract, registry, AnthropicTransport methods). 359 anthropic-related tests pass (0 failures). PR 3 of the provider transport refactor.
This commit is contained in:
kshitijk4poor
Teknium
parent
04f9ffb792
commit
731f4fbae6
5 changed files with 539 additions and 45 deletions
106
run_agent.py
106
run_agent.py
|
|
@ -6545,6 +6545,15 @@ class AIAgent:
|
|||
return suffix
|
||||
return "[A multimodal message was converted to text for Anthropic compatibility.]"
|
||||
|
||||
def _get_anthropic_transport(self):
|
||||
"""Return the cached AnthropicTransport instance (lazy singleton)."""
|
||||
t = getattr(self, "_anthropic_transport", None)
|
||||
if t is None:
|
||||
from agent.transports import get_transport
|
||||
t = get_transport("anthropic_messages")
|
||||
self._anthropic_transport = t
|
||||
return t
|
||||
|
||||
def _prepare_anthropic_messages_for_api(self, api_messages: list) -> list:
|
||||
if not any(
|
||||
isinstance(msg, dict) and self._content_has_image_parts(msg.get("content"))
|
||||
|
|
@ -6661,20 +6670,14 @@ 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
|
||||
_transport = self._get_anthropic_transport()
|
||||
anthropic_messages = self._prepare_anthropic_messages_for_api(api_messages)
|
||||
# Pass context_length (total input+output window) so the adapter can
|
||||
# clamp max_tokens (output cap) when the user configured a smaller
|
||||
# context window than the model's native output limit.
|
||||
ctx_len = getattr(self, "context_compressor", None)
|
||||
ctx_len = ctx_len.context_length if ctx_len else None
|
||||
# _ephemeral_max_output_tokens is set for one call when the API
|
||||
# returns "max_tokens too large given prompt" — it caps output to
|
||||
# the available window space without touching context_length.
|
||||
ephemeral_out = getattr(self, "_ephemeral_max_output_tokens", None)
|
||||
if ephemeral_out is not None:
|
||||
self._ephemeral_max_output_tokens = None # consume immediately
|
||||
return build_anthropic_kwargs(
|
||||
return _transport.build_kwargs(
|
||||
model=self.model,
|
||||
messages=anthropic_messages,
|
||||
tools=self.tools,
|
||||
|
|
@ -7356,9 +7359,9 @@ class AIAgent:
|
|||
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(
|
||||
# Native Anthropic — use the transport for kwargs
|
||||
_tflush = self._get_anthropic_transport()
|
||||
ant_kwargs = _tflush.build_kwargs(
|
||||
model=self.model, messages=api_messages,
|
||||
tools=[memory_tool_def], max_tokens=5120,
|
||||
reasoning_config=None,
|
||||
|
|
@ -7386,10 +7389,15 @@ class AIAgent:
|
|||
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, strip_tool_prefix=self._is_anthropic_oauth)
|
||||
if _flush_msg and _flush_msg.tool_calls:
|
||||
tool_calls = _flush_msg.tool_calls
|
||||
_tfn = self._get_anthropic_transport()
|
||||
_flush_nr = _tfn.normalize_response(response, strip_tool_prefix=self._is_anthropic_oauth)
|
||||
if _flush_nr and _flush_nr.tool_calls:
|
||||
tool_calls = [
|
||||
SimpleNamespace(
|
||||
id=tc.id, type="function",
|
||||
function=SimpleNamespace(name=tc.name, arguments=tc.arguments),
|
||||
) for tc in _flush_nr.tool_calls
|
||||
]
|
||||
elif hasattr(response, "choices") and response.choices:
|
||||
assistant_message = response.choices[0].message
|
||||
if assistant_message.tool_calls:
|
||||
|
|
@ -8449,14 +8457,14 @@ class AIAgent:
|
|||
summary_kwargs["extra_body"] = summary_extra_body
|
||||
|
||||
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,
|
||||
_tsum = self._get_anthropic_transport()
|
||||
_ant_kw = _tsum.build_kwargs(model=self.model, messages=api_messages, tools=None,
|
||||
max_tokens=self.max_tokens, reasoning_config=self.reasoning_config,
|
||||
is_oauth=self._is_anthropic_oauth,
|
||||
preserve_dots=self._anthropic_preserve_dots())
|
||||
summary_response = self._anthropic_messages_create(_ant_kw)
|
||||
_msg, _ = _nar(summary_response, strip_tool_prefix=self._is_anthropic_oauth)
|
||||
final_response = (_msg.content or "").strip()
|
||||
_sum_nr = _tsum.normalize_response(summary_response, strip_tool_prefix=self._is_anthropic_oauth)
|
||||
final_response = (_sum_nr.content or "").strip()
|
||||
else:
|
||||
summary_response = self._ensure_primary_openai_client(reason="iteration_limit_summary").chat.completions.create(**summary_kwargs)
|
||||
|
||||
|
|
@ -8481,14 +8489,14 @@ class AIAgent:
|
|||
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,
|
||||
_tretry = self._get_anthropic_transport()
|
||||
_ant_kw2 = _tretry.build_kwargs(model=self.model, messages=api_messages, tools=None,
|
||||
is_oauth=self._is_anthropic_oauth,
|
||||
max_tokens=self.max_tokens, reasoning_config=self.reasoning_config,
|
||||
preserve_dots=self._anthropic_preserve_dots())
|
||||
retry_response = self._anthropic_messages_create(_ant_kw2)
|
||||
_retry_msg, _ = _nar2(retry_response, strip_tool_prefix=self._is_anthropic_oauth)
|
||||
final_response = (_retry_msg.content or "").strip()
|
||||
_retry_nr = _tretry.normalize_response(retry_response, strip_tool_prefix=self._is_anthropic_oauth)
|
||||
final_response = (_retry_nr.content or "").strip()
|
||||
else:
|
||||
summary_kwargs = {
|
||||
"model": self.model,
|
||||
|
|
@ -9357,16 +9365,13 @@ class AIAgent:
|
|||
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:
|
||||
_tv = self._get_anthropic_transport()
|
||||
if not _tv.validate_response(response):
|
||||
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 not content_blocks:
|
||||
response_invalid = True
|
||||
error_details.append("response.content is empty")
|
||||
if response is None:
|
||||
error_details.append("response is None")
|
||||
else:
|
||||
error_details.append("response.content invalid (not a non-empty list)")
|
||||
else:
|
||||
if response is None or not hasattr(response, 'choices') or response.choices is None or not response.choices:
|
||||
response_invalid = True
|
||||
|
|
@ -9527,8 +9532,8 @@ class AIAgent:
|
|||
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")
|
||||
_tfr = self._get_anthropic_transport()
|
||||
finish_reason = _tfr.map_finish_reason(response.stop_reason)
|
||||
else:
|
||||
finish_reason = response.choices[0].finish_reason
|
||||
assistant_message = response.choices[0].message
|
||||
|
|
@ -9557,10 +9562,24 @@ class AIAgent:
|
|||
if self.api_mode in ("chat_completions", "bedrock_converse"):
|
||||
_trunc_msg = response.choices[0].message if (hasattr(response, "choices") and response.choices) else None
|
||||
elif self.api_mode == "anthropic_messages":
|
||||
from agent.anthropic_adapter import normalize_anthropic_response
|
||||
_trunc_msg, _ = normalize_anthropic_response(
|
||||
_trunc_nr = self._get_anthropic_transport().normalize_response(
|
||||
response, strip_tool_prefix=self._is_anthropic_oauth
|
||||
)
|
||||
_trunc_msg = SimpleNamespace(
|
||||
content=_trunc_nr.content,
|
||||
tool_calls=[
|
||||
SimpleNamespace(
|
||||
id=tc.id, type="function",
|
||||
function=SimpleNamespace(name=tc.name, arguments=tc.arguments),
|
||||
) for tc in (_trunc_nr.tool_calls or [])
|
||||
] or None,
|
||||
reasoning=_trunc_nr.reasoning,
|
||||
reasoning_content=None,
|
||||
reasoning_details=(
|
||||
_trunc_nr.provider_data.get("reasoning_details")
|
||||
if _trunc_nr.provider_data else None
|
||||
),
|
||||
)
|
||||
|
||||
_trunc_content = getattr(_trunc_msg, "content", None) if _trunc_msg else None
|
||||
_trunc_has_tool_calls = bool(getattr(_trunc_msg, "tool_calls", None)) if _trunc_msg else False
|
||||
|
|
@ -9819,9 +9838,10 @@ class AIAgent:
|
|||
# Log cache hit stats when prompt caching is active
|
||||
if self._use_prompt_caching:
|
||||
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
|
||||
_tcs = self._get_anthropic_transport()
|
||||
_cache = _tcs.extract_cache_stats(response)
|
||||
cached = _cache["cached_tokens"] if _cache else 0
|
||||
written = _cache["creation_tokens"] if _cache else 0
|
||||
else:
|
||||
# OpenRouter uses prompt_tokens_details.cached_tokens
|
||||
details = getattr(response.usage, 'prompt_tokens_details', None)
|
||||
|
|
@ -10766,15 +10786,13 @@ class AIAgent:
|
|||
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_v2
|
||||
_nr = normalize_anthropic_response_v2(
|
||||
_transport = self._get_anthropic_transport()
|
||||
_nr = _transport.normalize_response(
|
||||
response, strip_tool_prefix=self._is_anthropic_oauth
|
||||
)
|
||||
# Back-compat shim: downstream code expects SimpleNamespace with
|
||||
# .content, .tool_calls, .reasoning, .reasoning_content,
|
||||
# .reasoning_details attributes. This shim makes the cost of the
|
||||
# old interface visible — it vanishes when the full transport
|
||||
# wiring lands (PR 3+).
|
||||
# .reasoning_details attributes.
|
||||
assistant_message = SimpleNamespace(
|
||||
content=_nr.content,
|
||||
tool_calls=[
|
||||
|
|
|
|||
Loading…
Add table
Add a link
Reference in a new issue