cross-channel-history-retrieval

Name: cross-channel-history-retrieval
Availability: InStock
Author: openJiuwen

Search local history.json files across sessions and channels for matching messages.

2.5k|422|Updated Mar 5, 2026

Search Tags:#chat history #keyword matching #context injection #cross-channel #memory retrieval #agent sessions #time window search

Ready to run

npx skills add https://github.com/openJiuwen-ai/jiuwenswarm --skill cross-channel-history-retrieval-openjiuwen-ai

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: cross-channel-history-retrieval
Source: https://github.com/openJiuwen-ai/jiuwenswarm/tree/main/jiuwenswarm/resources/agent/workspace/skills/cross-channel-history-retrieval
Command: npx skills add https://github.com/openJiuwen-ai/jiuwenswarm --skill cross-channel-history-retrieval-openjiuwen-ai

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

When you ask about past events, people, dates, or earlier conversations but the current memory is insufficient, this Skill fetches the original chat content from other sessions/channels to ground your answer.

Core Features & Use Cases

Cross-session history search: Reads history.json from ~/.jiuwenswarm/agent/sessions/*/history.json to find matching messages across different channels and sessions.
Time-window and keyword filtering: Supports --channel, --session-id, keyword matching, and flexible time ranges using --start/--end or --at with --window-minutes.
Auto-expansion on no hits: If nothing is found and auto-expansion is enabled, it retries with a larger time window to improve recall.
Context-ready output: Produces a summary plus a HISTORY_CONTEXT_BLOCK that can be pasted directly into the active conversation context.

Quick Start

Ask the AI to run cross-channel-history-retrieval with your target channel (or no channel), keywords, and a time window so it returns the matching original messages to use in your current answer.

Frequently Asked Questions about cross-channel-history-retrieval

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I retrieve missing chat history from other agent sessions?▼

To retrieve missing chat history, this Skill searches local history.json files across different agent sessions and channels, filtering by keywords and time windows to return matching original messages directly into your active conversation context.

Can I search for specific conversation messages across multiple channels at once?▼

Yes, you can search conversation messages across multiple channels by specifying a target channel or leaving it open, applying keyword matching and flexible time ranges to scan local session histories for relevant cross-channel dialogue.

What happens if a cross-channel history search returns no matching messages?▼

If a cross-channel history search finds no matching messages and auto-expansion is enabled, the Skill automatically retries using a larger time window to improve recall and locate the missing conversation content.

How do I filter chat history by a specific time range and keywords?▼

You filter chat history by providing target keywords alongside time boundaries using start and end flags, or by specifying an exact time with a window in minutes, allowing the script to scan and match historical messages precisely.

Do I need any external dependencies to search local chat history?▼

No external dependencies are required to search local chat history; the Skill executes a provided Python script via mcp_exec_command to scan existing local history.json files stored within your agent session directories.

Why does my current agent memory fail to recall past conversations?▼

Current agent memory fails to recall past conversations when the context is insufficient, requiring this Skill to fetch original chat content from other sessions to ground answers about past events, people, and dates.