========== YENİ KOMUT | 23.09.2026 02:01:47 ==========
/home/hermes/.hermes/hermes-agent/tests/plugins/test_memory_hook_registration.py:48: return manager.invoke_hook("pre_llm_call", session_id="")
/home/hermes/.hermes/hermes-agent/tests/plugins/test_langfuse_plugin.py:161: mod.on_pre_llm_call(task_id="t", session_id="s", messages=[{"role": "user", "content": "hi"}])
/home/hermes/.hermes/hermes-agent/tests/hermes_cli/test_plugins.py:1069: assert mgr.invoke_hook("pre_llm_call", session_id="s1") == [
/home/hermes/.hermes/hermes-agent/tests/hermes_cli/test_plugins.py:1969: "pre_llm_call", session_id="s1", user_message="hi",
/home/hermes/.hermes/hermes-agent/tests/hermes_cli/test_plugins.py:2002: "pre_llm_call", session_id="s1", user_message="hi",
/home/hermes/.hermes/hermes-agent/plugins/observability/langfuse/__init__.py:675:def on_pre_llm_call(*, task_id: str = "", session_id: str = "", platform: str = "", model: str = "",
/home/hermes/.hermes/hermes-agent/docs/observability/README.md:114:Common `pre_llm_call` fields include `session_id`, `turn_id`,
/home/hermes/.hermes/hermes-agent/website/docs/user-guide/features/hooks.md:446:| `pre_llm_call` | Directive/control | Once per turn before the loop; all valid string/`{"context": ...}` returns are joined and injected into the user message. | `session_id`, `task_id`, `turn_id`, `user_message`, `conversation_history`, `is_first_turn`, `model`, `platform`, `parent_session_id`, `sender_id` | Full user message and conversation history. |
/home/hermes/.hermes/hermes-agent/website/docs/developer-guide/plugins/index.md:945:| [`pre_llm_call`](/user-guide/features/hooks#pre_llm_call) | Once per turn, before the tool-calling loop | `session_id: str, user_message: str, conversation_history: list, is_first_turn: bool, model: str, platform: str` | [context injection](#pre_llm_call-context-injection) |
/home/hermes/.hermes/hermes-agent/website/docs/developer-guide/plugins/index.md:965:The **API request hooks** are observers for the raw provider request, one level below the per-turn `pre_llm_call` / `post_llm_call` pair: a single turn that calls tools makes several API requests, and these hooks fire around each one. They exist for observability plugins (tracing, cost accounting, latency dashboards). The `request` and `response` kwargs are sanitized, size-capped JSON views of the provider payload (sensitive keys redacted, long strings truncated, SDK objects normalized), and `usage` is a plain token-summary dict. Every payload carries the correlation fields `turn_id`, `api_request_id`, `task_id`, `session_id`, and `api_call_count`, so a plugin can stitch requests, tool calls, and turns together. `api_request_error` fires when a provider call raises and adds `status_code`, `retry_count` / `max_retries`, `retryable`, `reason`, and an `error` dict with `type` and `message`.
/home/hermes/.hermes/hermes-agent/website/i18n/zh-Hans/docusaurus-plugin-content-docs/current/user-guide/features/hooks.md:385:| `pre_llm_call` | 指令/控制 | 每轮 loop 前一次;所有有效字符串或 `{"context": ...}` 会拼接并注入用户消息。 | `session_id`, `task_id`, `turn_id`, `user_message`, `conversation_history`, `is_first_turn`, `model`, `platform`, `parent_session_id`, `sender_id` | 完整用户消息和会话历史。 |
/home/hermes/.hermes/hermes-agent/website/i18n/zh-Hans/docusaurus-plugin-content-docs/current/developer-guide/plugins/index.md:536:| [`pre_llm_call`](/user-guide/features/hooks#pre_llm_call) | 每轮一次,工具调用循环前 | `session_id: str, user_message: str, conversation_history: list, is_first_turn: bool, model: str, platform: str` | [上下文注入](#pre_llm_call-context-injection) |
root@213-238-170-219:/home/hermes#