tommy0103/obelisk

Every past session, subagent, and workflow -- queryable by your agent, browsable by you

What it solves

Obelisk provides a unified, searchable index of session histories from multiple AI coding agents (such as Claude Code, Codex, DeepSeek Harness, Kimi Code, OMP, and Pi). It solves the problem of fragmented session logs—often stored as scattered JSONL files—by consolidating them into a single SQLite database, making them queryable for both humans and AI agents.

How it works

Obelisk uses a CLI and a desktop app to index transcripts from various provider-specific directories. It projects these diverse formats into a common SQLite schema using provider adapters.

  • Agent Side: A CLI tool allows coding agents to write and execute JavaScript queries against the index to retrieve specific information from past sessions.
  • App Side: An Electron-based desktop application allows humans to browse sessions, view tool calls, and track usage statistics.
  • Memory Layer: Agents can propose markdown memory files based on retrieval results, which are then registered in the index as a synthesis cache for future recall.

Who it’s for

Developers using multiple AI coding agents and those who want their agents to have a long-term, cross-session memory of their project's evolution and technical decisions.

Highlights

  • Multi-provider support: Unified indexing for Claude Code, Codex, Kimi Code, OMP, Pi, and DeepSeek Harness.
  • Agent-first retrieval: Enables agents to use a /obelisk skill to search history and answer questions in natural language.
  • Human-readable surface: A desktop app for browsing sessions, tool calls, and viewing activity heatmaps.
  • Incremental indexing: Only new or modified files are re-parsed to maintain performance.
  • Memory synthesis: A mechanism for agents to create and register persistent markdown memories.

Related

  • Project
    AVIDS2/memorix

    Memorix is a Node.js library that adds a local‑first, searchable memory store to any AI coding assistant. It persists facts, design rationales, and Git‑derived insights in a SQLite database, exposing them via MCP, a CLI, and an optional web UI. By installing a single `memorix setup --agent <name>` command, dozens of agents (Claude Code, Codex, Copilot CLI, Cursor, Gemini CLI, etc.) can share the same project‑scoped memory, enabling smoother handoffs, better context retention, and multi‑agent orchestration.

  • Project
    nvk/llm-wiki

    llm‑wiki is a cross‑runtime tool that lets Claude, Codex, OpenCode, Pi and other LLM agents build and query a structured, Obsidian‑compatible wiki. It provides ingestion, parallel research, session memory, private adapters, and exportable knowledge checkpoints, with install scripts for each supported agent.

  • Project
    codeaholicguy/ai-devkit

    AI DevKit is a local npm‑based control plane that unifies configuration, monitoring, cross‑agent messaging, searchable SQLite memory, and composable engineering “skills” (requirements, design, testing, review, etc.) for many AI coding assistants such as Claude Code, Gemini CLI, Codex CLI, Cursor, and GitHub Copilot. It lets teams manage multiple agents from a single console, keep project conventions in a searchable store, and enforce a senior‑engineer workflow without modifying the underlying LLMs.

  • Project
    volcengine/OpenViking

    OpenViking is an open‑source context database that presents an LLM agent’s memories, resources and skills as a virtual filesystem (`viking://`). Each entry is pre‑processed into three layers (abstract, overview, full detail) so the agent only loads what it needs, dramatically reducing token usage and latency while keeping retrieval paths observable. It ships with a Python server, a CLI (`ov`), a web Studio demo, and ready‑made integrations for many popular agents (Claude Code, Codex, LangChain, etc.). Install via `pip install openviking`, run `openviking-server init` to configure providers, then start the service and use `ov` commands to add resources, browse, and search. Benchmarks show 30‑90 % token savings and accuracy jumps to ~80 % on memory‑heavy tasks. The core is AGPL‑v3; a managed SaaS version is also offered by Volcano Engine.

  • Project
    HoangNguyen0403/agent-skills-standard

    Agent Skills Standard (AGS) is an NPM‑distributed CLI that lets teams publish their coding‑standards, security rules, and SDLC workflow templates as tiny markdown “skills”. The CLI (`ags init` / `ags sync`) creates a `.skillsrc` config, downloads language‑specific skill packs (280 skills across 20+ stacks), and writes a hierarchical router (`AGENTS.md → _INDEX.md → SKILL.md`). AI coding assistants (Cursor, Claude Code, Copilot, Gemini, etc.) read this router and load only the relevant rules for the file they are editing, saving ~85 % of prompt tokens. An optional MCP server can enforce hooks, block secret‑leak edits, and provide runtime audit logs. The project is a genuine AI‑coding‑assistant utility, not a generic tutorial or list.