Skip to content
TheChatArchiveThe Chat Archive
composer-2.5· unknownHugging Face ibragim-bad/swe_rebench_07_2026_trajectories

history.jsonl 跨会话注入导致上下文污染 Bug Description history.jsonl 跨会话注入导致上下文污染 问题描述 history.jsonl…

  1. User

    history.jsonl 跨会话注入导致上下文污染

    Bug Description

    history.jsonl 跨会话注入导致上下文污染

    问题描述

    history.jsonl 中的每条 entry 来源于 单个会话 的消息摘要(通过 Consolidator.archive() 生成),但 ContextBuilder.build_system_prompt() 在注入 # Recent History 时,未做会话隔离,会将所有会话尚未被 Dream 处理的 entry 全部混入当前会话的 system prompt。

    数据流分析

    写入路径(Consolidator)

    Consolidator.archive() 被三个调用方触发,均传入 per-session 的消息块:

    # memory.py:578, 751, 824
    chunk = session.messages[session.last_consolidated:end_idx]  # 单会话切片
    summary = await self.archive(chunk)
    self.store.append_history(summary, max_chars=_ARCHIVE_SUMMARY_MAX_CHARS)
    

    每条写入的 JSONL entry 格式为:

    {"cursor": N, "timestamp": "2026-06-09 12:00", "content": "用户偏好深色主题..."}
    

    注意:entry 中没有 session_key 或 chat_id 字段,无法追溯来源会话。

    注入路径(ContextBuilder)

    # context.py:64-71
    entries = self.memory.read_unprocessed_history(since_cursor=self.memory.get_last_dream_cursor())
    if entries:
        capped = entries[-self._MAX_RECENT_HISTORY:]
        history_text = "\n".join(
            f"- [{e['timestamp']}] {e['content']}" for e in capped
        )
        history_text = truncate_text(history_text, self._MAX_HISTORY_CHARS)
        parts.append("# Recent History\n\n" + history_text)
    

    read_unprocessed_history(since_cursor) 读取的是 cursor > dream_cursor 的全部条目,不区分来源会话,无任何 session 隔离。

    问题场景

    时序事件dream_cursor
    T1微信会话 A 触发 Consolidator,写入 2 条摘要 entry0
    T2Telegram 会话 B 触发 Consolidator,写入 1 条摘要 entry0
    T3用户打开 Telegram 会话 B,调用 build_system_prompt读取 cursor>0 的 全部 3 条,混入会话 B 的 system prompt

    此时会话 B 的 # Recent History 中包含了微信会话 A 的摘要,而该摘要是从微信会话提炼的,可能包含:

    • 微信特定的事实或上下文
    • 与 Telegram 会话完全无关的用户偏好
    • 尚未被 Dream 消化/验证的原始摘要

    影响评估

    • 会话上下文污染 — 当前会话的 system prompt 被注入无关会话的事实,可能干扰 LLM 判断

    修复方向建议

    在 entry 中记录 session_key

    在 append_history() 的 entry 中增加 session_key 字段:

    {"cursor": N, "timestamp": "2026-06-09 12:00", "session_key": "telegram:12345", "content": "用户偏好深色主题..."}
    

    然后在 read_unprocessed_history() 或 ContextBuilder 注入时支持按 session_key 过滤。

    Steps to Reproduce

    无

    Expected Behavior

    无

    Relevant Logs

    nanobot Version

    0.2.0

    Python Version

    3.11

    Operating System

    macOS

    Channel / Platform

    Weixin (Personal WeChat)

    LLM Provider

    OpenAI

    Configuration (Optional)

    Additional Context

    No response

    Evaluation contract

    Every archive path—including summary, raw fallback, idle compaction, and /new archival—must propagate the originating session_key into history entries. Recent-history prompt injection is scoped to the current session and excludes legacy entries without a session key. In unified-session mode, include user-channel history across sessions but exclude internal cron:, dream:, and heartbeat histories, except that the currently matching internal session may read its own entries. Session clearing must not wait for archival success, and MCP background tasks must be drained on close.

    Interface

    Type: Function Name: append_history Path: nanobot/agent/memory.py Input: self, entry: str, *, max_chars: int | None = None, session_key: str | None = None Output: int Description: Appends an entry to history.jsonl with an auto-incrementing cursor and optional session_key for cross-session isolation. Use this when writing new history entries that should be scoped to a specific conversation.

    Type: Function Name: archive Path: nanobot/agent/memory.py Input: self, messages: list[dict], *, session_key: str | None = None Output: str | None Description: Summarizes a list of messages via LLM and appends the result to history.jsonl with the given session_key. Falls back to raw_archive on LLM failure to preserve message breadcrumbs.

    Type: Function Name: raw_archive Path: nanobot/agent/memory.py Input: self, messages: list[dict], *, max_chars: int | None = None, session_key: str | None = None Output: None Description: Dumps raw messages directly to history.jsonl without LLM summarization, tagged with an optional session_key. Used as a fallback when the LLM call in archive() fails.

    Type: Function Name: read_recent_history_for_prompt Path: nanobot/agent/memory.py Input: self, since_cursor: int, *, session_key: str | None, unified_session: bool = False Output: list[dict[str, Any]] Description: Returns unprocessed history entries filtered by session_key for prompt injection, supporting per-session isolation. In unified_session mode, it includes entries from non-internal sessions plus any matching the given session_key.

  2. Assistant

    Investigating how history is written and injected so we can add session isolation per the evaluation contract.

  3. Assistant

    Implementing session_key propagation across archive paths and scoped history injection.

  4. Assistant

    Updating failing tests and adding session-isolation coverage.

  5. Assistant

    Fixed cross-session context pollution in history.jsonl by tagging entries with session_key and scoping # Recent History injection to the current session.

    Changes

    MemoryStore (nanobot/agent/memory.py)

    • append_history() — optional session_key is persisted on each JSONL entry
    • raw_archive() — forwards session_key to append_history()
    • read_recent_history_for_prompt() — new filter for prompt injection:
      • Normal mode: only entries matching the current session_key
      • Unified-session mode: user-channel history across sessions; excludes cron:, dream:, and heartbeat unless the current session is that internal session
      • Legacy entries without session_key are excluded

    Consolidator (nanobot/agent/memory.py)

    • archive() — accepts session_key and propagates it through summary and raw-fallback paths
    • All consolidation call sites (replay overflow, token budget, idle compact) pass session.key

    ContextBuilder (nanobot/agent/context.py)

    • build_system_prompt() / build_messages() — accept session_key and unified_session
    • Recent history now uses read_recent_history_for_prompt() instead of unfiltered read_unprocessed_history()

    AgentLoop (nanobot/agent/loop.py)

    • Passes session_key and unified_session into message building
    • enforce_file_cap uses partial(raw_archive, session_key=...) so file-cap archives are tagged

    /new command (nanobot/command/builtin.py)

    • Background archive passes session_key=ctx.key (session still clears immediately; archive remains fire-and-forget)

    Behavior

    ScenarioBeforeAfter
    WeChat archives → Telegram opensAll 3 entries injectedOnly Telegram entries
    Legacy entries (no session_key)IncludedExcluded from prompts
    Unified session modeN/ACross-channel user history; internal sessions filtered
    Dream processingUnchangedStill reads all unprocessed entries globally

    1110 agent tests pass, including new session-isolation coverage in test_memory_store.py.

Responses were produced in different, uncontrolled contexts (system prompts, settings, prior turns). This is not a benchmark.