Codex traces preview
Fri, Jul 31
# Files mentioned by the user: ## claude-opus-4.7-traceInversion-5000x.jsonl: C:\Users\crazy\Downloads\claude-opus-4.7-traceInversion-5000x.jsonl ## My request for Codex: 이거 messages chatml만 남기고, thinking 칼럼을 reasoning으로 바꿔줘. 그리고 파일로 저장해줘. 1 message 3 tools
Thu, Jul 30
여기 index.html을 수정해서 apache2로 호스팅 가능하게 production급으로 수정해줘. 그리고 다크모드 만들고, OS 기본값에 따라 바뀌게, 버튼으로 꺼짐/켜짐 가능하게, 쿠키로 저장하게 해줘. 4 messages 7 tools
Tue, Jul 28
이미지에서 기울어진 사각형의 4개의 포인트를 찍으면 그걸 원근으로 복원해주는 프로그램을 만들자. C로 만들어서, usb에 들고다닐수있는 .exe 파일로 컴파일하는거지. 3 messages 136 tools
Mon, Jul 27
이 폴더에 있는 원본 내보내기 파일을, 여러 세션의 멀티턴 chatml jsonl(messages, [{role, content, reasoning}])로 내보내는 프로그램을 작성해줘. 이때 이미지가 포함된거는 삭제, librechat의 형식으로 변환, librechat으로 변환 못하는건 삭제. librechat에 검색 기능 있음. 4 messages 25 tools
Sun, Jul 26
이 폴더에 있는 json 파일을, 여러 세션의 멀티턴 chatml jsonl(messages, [{role, content, reasoning}])로 내보내는 프로그램을 작성해줘. 3 messages 1 tool
Mon, Jul 20
여기서 데이터 머징할때 오류 안고치고 머징하면 멀쩡한것만 머징하는걸로 수정해줘. 6 messages 2 tools 현재 이 코드에서 Reasoning은 영어로. 질문이랑 출력만 한국어로 수정해줘. 그리고 기존 체크포인트 초기화하는 방법은? 1 message 이 코드에서 여러 백엔드 주소 지정해서 사용 가능하게 구현해줘. V100 16GB 1장, 3090 1장 있는데 각 GPU에 할당할 모델 추천해줘. VRAM 되면 여러 모델 같이 로딩도 가능. llama.cpp 쓸거임. 목적은 영어 --> 한국어 번역. 3 messages 순수 텍스트 1.45GB 양의 영어 Instruct를 한국어로 번역할 예정. hf datasets에서 chatml jsonl로 받은 데이터셋을 openai 호환 api를 통해 한국어로 번역, 이후 파싱후, 이상 있으면 다시, 8번 시도해서 안되면 오류 리스트에 등재, 오류 리스트는 사람이 고침. 이후 오류 리스트를 원본에 머지. 각 데이터 세션은 해시로 고유번호 발급, 체크포인트 관리해서, 중간에 끊어져도 다시 재개하면 잘못된 데이터, 번역 안한 데이터만 재개하게. 번역은 메시지 단위로 쪼개지 말고, 한 세션(Row) 전체를 통째로 해야해. 모델은 자체모델(Lumen, Gemini 3 기반) 2B, 4B, 12B, 26B, 그리고 GPT-OSS 20B 있는데 뭐해? 아니면 걍 OpenAI API 쓸까? 쓰면 어떤 모델, 얼마나 들어? 4 messages 2 tools
Sun, Jul 19
이 앱을 원본 그대로 폴더 이름과 클래스 이름만 추적 후 이후 개발이 순조롭게 이름 변경, 다시 재컴파일 가능하게 apk, 가능하면 윈도우& 맥& 리눅스도 빌드 스크립트 작성해줘. 그리고 블로트웨어, 광고 시스템, 암호화폐 채굴, 바이러스 등은 README.md에 섹션 뽑아서 기록만 해놔. 절대 건들지 말고 원본 유지가 목표. 이 프로젝트는 보안 연구임. 한때 유명했던 SNS 앱이 어떤 블로트웨어가 있는지, 어떤 바이러스가 있는지, 왜 이렇게 비정상적인 전력& 메모리 사용량이 있는지를 분석하는 연구. README.md도 너가 이에 대해 컴파일하지 말고 소스코드만 분석해서 연구, 작성해. 어떤 바이러스/블로트웨어/백도어 가 얼마나 있었다 이런 통계도 내고. 11 messages 58 tools

This dataset was generated using teich by TeichAI

My Agent Traces

This directory contains raw agent trace files generated by teich.

JSONL files: 20

Training-ready tools

Generated agent traces carry configured or recovered tool schemas so tools remain available for training even when a session did not call them. Native Claude Code imports recover schemas for Claude Code and Claude Desktop built-ins, plus conservative name-derived MCP schemas, when the raw transcript only records tool names or calls. A complete dataset-level tools schema snapshot is embedded in the collapsed section at the bottom of this README. load_traces applies the dataset snapshot to each loaded example as a fallback tools field.

Create A Similar Dataset

This dataset was staged from existing local agent sessions with teich extract. To build your own local dataset, install Teich and point it at one of the supported providers:

teich extract codex --out data

Use --sessions-dir /path/to/store when your sessions are not in the default location, and --model <substring> when you only want sessions whose model metadata matches a value.

Format

Each file is newline-delimited JSON representing a single captured agent session. The trace schema is designed for upload-first preservation so you can keep the original session history and convert it later for training. Teich normalizes split assistant fragments during trace copy and conversion so the semantic order is reasoning first, optional assistant text second, and tool calls last. Native Claude Code conversion also preserves runtime context such as skills, MCP instructions, hook context, permission state, date changes, and session recaps as masked system messages when the raw transcript provides them.

Common top-level event groups:

  • session_meta
  • turn_context
  • event_msg
  • response_item
  • session
  • message
  • session_info
  • model_change
  • thinking_level_change
  • external_session_meta
  • external_message
  • external_stderr

Example

{"timestamp": "2026-07-19T08:53:37.517Z", "type": "session_meta", "payload": {"session_id": "019f7994-b5ce-71d0-9fbc-1b66f4e1c7bd", "id": "019f7994-b5ce-71d0-9fbc-1b66f4e1c7bd", "timestamp": "2026-07-19T08:53:36.846Z", "cwd": "C:\\Users\\crazy\\OMG-Source", "originator": "Codex Desktop", "cli_version": "0.144.2", "__truncated__": "7 keys omitted"}}

Training

Use this dataset as Aeonthic/Agentic-SFT with Teich's data preparation and training utilities. If you do not want Teich to handle chat-template formatting or masking, run teich convert to write standalone OpenAI-style JSONL rows with prompt, messages, tools, and metadata. Training setup details evolve over time, so the maintained guide lives in the Teich training docs. For loading, mixing, converting, and validating Teich datasets, see Preparing Data.

Tool schema snapshot

Training-ready tool schema snapshot
[
  {
    "type": "function",
    "function": {
      "name": "apply_patch",
      "description": "Apply a unified patch to files in the workspace.",
      "parameters": {
        "type": "object",
        "properties": {
          "patch": {
            "type": "string"
          }
        },
        "additionalProperties": true,
        "required": [
          "patch"
        ]
      }
    }
  },
  {
    "type": "function",
    "function": {
      "name": "bash",
      "description": "Run shell commands in the workspace.",
      "parameters": {
        "type": "object",
        "properties": {
          "command": {
            "type": "string"
          },
          "timeout_ms": {
            "type": "integer"
          },
          "workdir": {
            "type": "string"
          }
        },
        "additionalProperties": true,
        "required": [
          "command"
        ]
      }
    }
  },
  {
    "type": "function",
    "function": {
      "name": "exec",
      "parameters": {
        "type": "object",
        "properties": {
          "input": {
            "type": "string"
          }
        },
        "additionalProperties": true,
        "required": [
          "input"
        ]
      }
    }
  },
  {
    "type": "function",
    "function": {
      "name": "exec_command",
      "description": "Run a shell command in the workspace.",
      "parameters": {
        "type": "object",
        "properties": {
          "cmd": {
            "type": "string"
          },
          "justification": {
            "type": "string"
          },
          "login": {
            "type": "boolean"
          },
          "max_output_tokens": {
            "type": "integer"
          },
          "prefix_rule": {
            "type": "array"
          },
          "sandbox_permissions": {
            "type": "string"
          },
          "shell": {
            "type": "string"
          },
          "tty": {
            "type": "boolean"
          },
          "workdir": {
            "type": "string"
          },
          "yield_time_ms": {
            "type": "integer"
          }
        },
        "additionalProperties": true,
        "required": [
          "cmd"
        ]
      }
    }
  },
  {
    "type": "function",
    "function": {
      "name": "load_workspace_dependencies",
      "parameters": {
        "type": "object",
        "properties": {},
        "additionalProperties": true
      }
    }
  },
  {
    "type": "function",
    "function": {
      "name": "run",
      "parameters": {
        "type": "object",
        "properties": {
          "response_length": {
            "type": "string"
          },
          "search_query": {
            "type": "array",
            "items": {
              "type": "object",
              "properties": {
                "q": {
                  "type": "string"
                }
              },
              "additionalProperties": true,
              "required": [
                "q"
              ]
            }
          }
        },
        "additionalProperties": true,
        "required": [
          "response_length",
          "search_query"
        ]
      }
    }
  },
  {
    "type": "function",
    "function": {
      "name": "shell_command",
      "parameters": {
        "type": "object",
        "properties": {
          "command": {
            "type": "string"
          },
          "justification": {
            "type": "string"
          },
          "prefix_rule": {
            "type": "array",
            "items": {
              "type": "string"
            }
          },
          "sandbox_permissions": {
            "type": "string"
          },
          "timeout_ms": {
            "type": "integer"
          },
          "workdir": {
            "type": "string"
          }
        },
        "additionalProperties": true,
        "required": [
          "command",
          "timeout_ms",
          "workdir"
        ]
      }
    }
  },
  {
    "type": "function",
    "function": {
      "name": "update_plan",
      "description": "Update the current task plan.",
      "parameters": {
        "type": "object",
        "properties": {
          "explanation": {
            "type": "string"
          },
          "plan": {
            "type": "array"
          }
        },
        "additionalProperties": true,
        "required": [
          "plan"
        ]
      }
    }
  },
  {
    "type": "function",
    "function": {
      "name": "view_image",
      "description": "Inspect a local image file.",
      "parameters": {
        "type": "object",
        "properties": {
          "detail": {
            "type": "string"
          },
          "path": {
            "type": "string"
          }
        },
        "additionalProperties": true,
        "required": [
          "path"
        ]
      }
    }
  },
  {
    "type": "function",
    "function": {
      "name": "wait",
      "parameters": {
        "type": "object",
        "properties": {
          "cell_id": {
            "type": "string"
          },
          "max_tokens": {
            "type": "integer"
          },
          "terminate": {
            "type": "boolean"
          },
          "yield_time_ms": {
            "type": "integer"
          }
        },
        "additionalProperties": true,
        "required": [
          "cell_id",
          "yield_time_ms"
        ]
      }
    }
  }
]
Downloads last month
32