</>MCP Agents Market
Agent

coding_agent_session_search

by Dicklesworthstone1.1kRustUpdated 2026-08-24

Unified TUI and CLI to index and search your local coding agent session history across 11+ providers (Codex, Claude, Gemini, Cursor, Aider, etc.)

CodexClaude CodeCursorChatGPTAiderGemini CLIGitHub Copilot

coding_agent_session_search (cass) is a unified search interface that indexes and searches your local coding agent session history across 26+ providers including Codex, Claude Code, Gemini CLI, Cursor, Aider, ChatGPT, and more. It aggregates conversations from different AI coding assistants into a single searchable timeline with both lexical (BM25) and optional semantic search capabilities. The tool enables agents to discover past solutions across different agent sessions, preventing repeated problem-solving and creating a unified knowledge base from fragmented agent histories. It provides both an interactive TUI for humans and a JSON robot mode optimized for AI agent consumption.

Key Features

Indexes sessions from 26+ coding agents (Claude Code, Codex, Cursor, Gemini, Aider, ChatGPT, Copilot, etc.) into unified SQLite + search index
Sub-60ms search latency with BM25 lexical search via Quill; optional local semantic search using MiniLM embeddings (no network required)
Robot mode with structured JSON output, self-documenting API, and forgiving syntax for AI agent consumption
Cited answer packs (pack command) that bundle token-budgeted evidence with freshness and privacy metadata for agent handoffs
Multi-machine search via SSH/rsync to aggregate sessions from laptop, desktop, and remote servers with provenance tracking
Rich TUI with live search-as-you-type, multiple ranking modes, filtering, and analytics dashboard
Automatic background indexing with watch mode, scheduled jobs (launchd/systemd), and stale-on-read catch-up
HTML/Markdown export with optional AES-256-GCM encryption for shareable session archives

Use Cases

  • 01AI agents searching their own history and other agents' sessions for similar problems before solving from scratch
  • 02Cross-agent knowledge transfer: finding solutions discovered in Cursor while working in Claude Code
  • 03Debugging workflow archaeology: locating past debugging sessions for specific files or error patterns
  • 04Building training data by exporting high-quality problem-solving sessions across multiple agents
  • 05Multi-machine development: searching sessions from laptop, desktop, and remote servers in one unified index
  • 06Agent-to-agent handoff with cited evidence packs containing token-budgeted context and provenance metadata

Related Agents

View more

hermes-agent

The agent that grows with you

Agent250.4k

agency-agents

A complete AI agency at your fingertips - From frontend wizards to Reddit community ninjas, from whimsy injectors to reality checkers. Each agent is a specialized expert with personality, processes, and proven deliverables.

Agent155.5k

openinterpreter

A coding agent for open models like Kimi K3

Agent68k

cline

Autonomous coding agent as an SDK, IDE extension, or CLI assistant.

Agent66.3k

coding_agent_session_search — FAQ

What is coding_agent_session_search?+

It's a unified search tool that indexes and searches conversation history from 26+ local coding agents (Claude Code, Codex, Cursor, Aider, etc.) into a single searchable timeline. It provides both an interactive TUI and a JSON robot mode optimized for AI agent consumption.

How do I install cass?+

Install via Homebrew (brew install dicklesworthstone/tap/cass), Scoop on Windows, or the install script (curl -fsSL https://raw.githubusercontent.com/Dicklesworthstone/coding_agent_session_search/main/install.sh | bash -s -- --easy-mode --verify). After installation, run 'cass' to start the TUI or 'cass triage --json' for robot mode.

Which AI agents and clients does it work with?+

It indexes local session files from Codex, Claude Code, Cline, Gemini CLI, OpenCode, Amp, Cursor, ChatGPT, Aider, Pi-Agent, Oh My Pi, GitHub Copilot Chat, Copilot CLI, OpenClaw, Clawdbot, Vibe, Crush, Goose, Hermes, Kimi Code, Muse Code, Qwen Code, Factory, Antigravity, OpenHands, and Grok Build. It works as a standalone search tool that any agent can query via its robot mode API.

Do I need API keys or network access?+

No API keys are required. The tool reads local session files only. Semantic search uses local MiniLM model files (~90 MB, opt-in via 'cass models install') with no network calls at query time. The initial model download requires network access but can be done air-gapped with --from-file.

Is it free and open source?+

Yes, it's open source under MIT + OpenAI/Anthropic Rider license. The tool is free to use and all processing happens locally.

How do AI agents use this tool?+

Agents should use robot mode commands like 'cass triage --json' for preflight checks, 'cass search "query" --robot' for searches, and 'cass pack "query" --robot' for cited evidence bundles. Never run bare 'cass' in agent context as it launches the interactive TUI. The tool provides self-documenting capabilities via 'cass capabilities --json'.

How do I install coding_agent_session_search?+

Open the source repository on GitHub and follow its README. coding_agent_session_search is a agent — MCP Agents Market links you directly to the official repo.

Is coding_agent_session_search free?+

coding_agent_session_search is an open-source project hosted on GitHub. Check the repository for its license and any usage requirements.

Related searches

Developer Tools AgentsSearch AgentsKnowledge & Memory AgentsBest MCP ServersBest SkillsBest PluginsTop rated listingsClaude Code toolsBrowse all categories