{
  "schema": "tesign-radar/item/1",
  "generated_at": "2026-09-17T01:56:13.055Z",
  "lang": "en",
  "slug": "agent-memory",
  "url": "https://tesign.com/en/item/agent-memory/",
  "alternates": {
    "ko": {
      "html": "https://tesign.com/item/agent-memory/",
      "json": "https://tesign.com/item/agent-memory/index.json",
      "markdown": "https://tesign.com/item/agent-memory/index.md"
    },
    "en": {
      "html": "https://tesign.com/en/item/agent-memory/",
      "json": "https://tesign.com/en/item/agent-memory/index.json",
      "markdown": "https://tesign.com/en/item/agent-memory/index.md"
    }
  },
  "name": "agent-memory",
  "use_label": "Long-term memory for agent users",
  "one_liner": "Keeps what an agent learned as Markdown files it can search and reread.",
  "text": {
    "why": "Close a session and the agent forgets everything it learned there. agent-memory stores memories as Markdown files, ranks them with a SQLite index beside them, and answers with file paths rather than pasted text so the agent reads only as deep as needed. Writes fire at conversation boundaries, a separate sleep-time pass consolidates, and none of it needs an API key. Claude Code and Codex CLI share one store under MIT; no PyPI release yet, so it installs from a checkout.",
    "take": "It refuses to choose between a retrieval engine and a plain filesystem, and layers both on one store.",
    "build": [
      "A shared memory store where the agent finds last week's decisions again",
      "A setup where Claude Code and Codex draw on the same context",
      "A weekly routine that turns stale memories into proposals to confirm"
    ],
    "who": [
      {
        "role": "Daily agent user",
        "situation": "tired of re-explaining the same context each session"
      },
      {
        "role": "Two-CLI user",
        "situation": "wants Claude Code and Codex to share one memory"
      }
    ],
    "start": [
      {
        "command": true,
        "text": "git clone https://github.com/tigerless-labs/agent-memory.git"
      },
      {
        "command": true,
        "text": "cd agent-memory"
      },
      {
        "command": true,
        "text": "uv sync --all-packages"
      },
      {
        "command": true,
        "text": "mem init"
      },
      {
        "command": true,
        "text": "mem setup --host claude-code   # or: --host codex"
      },
      {
        "command": false,
        "text": "Needs Python 3.12+ and uv · put .venv/bin on PATH or the hook records nothing"
      }
    ],
    "caveat": [
      "No PyPI release yet; install from a checkout",
      "Version 0.1.0; the MCP surface is narrower than the CLI",
      "Judgement is borrowed from the host CLI, so a host agent is required"
    ]
  },
  "cta": {
    "type": "install",
    "label": "INSTALL",
    "url": "https://github.com/tigerless-labs/agent-memory"
  },
  "ready": "install",
  "categories": [
    "ai",
    "dev-tools"
  ],
  "categories_confirmed": true,
  "tags": [
    "agent-memory",
    "markdown",
    "local-first",
    "mcp"
  ],
  "platform": [
    "cli"
  ],
  "license": {
    "spdx": "MIT",
    "scope": "permissive",
    "scope_text": "Use, change and redistribute, commercially too. Keep the notice.",
    "url": "https://spdx.org/licenses/MIT.html",
    "note": null,
    "open_source": true
  },
  "language": "Python",
  "source_urls": {
    "canonical": "https://github.com/tigerless-labs/agent-memory",
    "github": "https://github.com/tigerless-labs/agent-memory",
    "huggingface": null,
    "app_store": null,
    "sightings": [
      {
        "source": "github",
        "url": "https://github.com/tigerless-labs/agent-memory",
        "posted_at": "2026-09-13T00:00:00.000Z",
        "hn_url": null
      }
    ]
  },
  "numbers": {
    "as_of": "2026-09-16T21:00:00.000Z",
    "stars": 1457,
    "stars_unit": "GitHub stars",
    "stars_checked_at": "2026-09-15T19:34:00.819Z",
    "stars_observed_since_check": 0,
    "star_delta_24h": null,
    "star_delta_7d": 109,
    "star_delta_30d": 260,
    "star_delta_unit": "GitHub stars gained in the window, GH Archive events summed to as_of",
    "spark_14d": [
      null,
      null,
      74,
      59,
      11,
      4,
      3,
      2,
      42,
      37,
      28,
      null,
      null,
      null
    ],
    "hf_likes": null,
    "hf_likes_observed_at": null,
    "hn_points": null,
    "hn_points_observed_at": null,
    "store": null,
    "method": "Star total = the value last checked on GitHub (stars_checked_at) + increases observed via GH Archive since. The 24h · 7d · 30d gains are GH Archive hourly events summed to the reference time (as_of). No score of ours."
  },
  "ranks": null,
  "activity": {
    "pushed_at": "2026-09-15T08:32:53.000Z",
    "owner_type": "Organization",
    "open_issues": 19,
    "contributors": 2,
    "releases_count": 0,
    "latest_release": null,
    "checked_at": "2026-09-15T19:34:00.819Z",
    "core_checked_at": "2026-09-15T19:34:00.819Z"
  },
  "signals": [],
  "images": {
    "og": "https://tesign.com/img/og/agent-memory.en.png",
    "cover": "https://tesign.com/img/agent-memory-f80a30a156.png",
    "cover_width": 1200,
    "cover_height": 600,
    "gallery": [
      "https://tesign.com/img/agent-memory-f80a30a156.png"
    ],
    "alt": "GitHub social card for tigerless-labs/agent-memory: name and description left, purple logo right.",
    "tier": 4
  },
  "dates": {
    "first_seen_at": "2026-09-13T07:43:30.193Z",
    "source_created_at": "2026-09-01T21:52:29.000Z",
    "discovery_hours": 273.8503313888889,
    "launch_signal_at": "2026-09-13T00:00:00.000Z",
    "kept_at": "2026-09-14T02:41:58.947Z",
    "first_published_at": null,
    "updated_at": "2026-09-15T04:36:56.786Z",
    "ingestion_mode": "live"
  },
  "notes": [
    "This file was produced by the same build, from the same data, as the tesign.com item page. The text is editorial; the numbers are stored observations.",
    "null in the JSON means not observed — never zero. The Markdown writes [unconfirmed] for it.",
    "Star total = the value last checked on GitHub (stars_checked_at) + increases observed via GH Archive since. The 24h · 7d · 30d gains are GH Archive hourly events summed to the reference time (as_of). No score of ours.",
    "A summary, not legal advice."
  ]
}