# Agendum

Provides persistent memory for AI agents, converts plans into task graphs, and scopes context across coding sessions.

- MCP server · `type=mcp` · Claude Code, Codex, Cursor, GitHub Copilot, Windsurf (Devin Desktop), Cline, Zed, Gemini CLI · ★ 6
- Author: sralli
- License: MIT · policy=link-only · description only; the file stays with the author
- Source: https://github.com/sralli/agendum
- Updated: 2026-04-13
- HTML: https://affpapa.org/en/nevibecode/mcp/agendum

## What it does

AI coding assistants lose project state between sessions. Agendum tackles this by storing task backlogs and architectural decisions inside the `.agendum/` directory. Everything lives in plain Markdown and YAML, making the entire project state diffable and readable in git. Running the server requires Python 3.13+ and `uvx`.

The server exposes 14 tools prefixed with `pm_`. Running `pm_ingest` parses a plan file and maps it into discrete tasks with dependency ordering. When invoked via `pm_next`, the agent receives an isolated work package with acceptance criteria and an adaptive context budget between 4K and 10K characters. Finishing work with `pm_done` records code patterns, reads `git diff`, and unblocks dependent items automatically.

For knowledge retrieval, `pm_search` relies on FTS5 text search and topic entity graphs. Semantic search is not active out of the box; it requires installing the `agendum[vectors]` extra. Built-in maintenance tools like `pm_consolidate` deduplicate findings and flag contradictory notes.

## Requirements
- Needs: Python and uv (commands run via uvx)

## Install

### Claude Code (native platform)

**Platform CLI**
```bash
claude mcp add agendum -- uvx agendum
```
uvx runs the command named after the package; if the README names another one, use uvx --from agendum <command>

### Codex

**Platform CLI**
```bash
codex mcp add agendum -- uvx agendum
```
uvx runs the command named after the package; if the README names another one, use uvx --from agendum <command>

### Cursor

**Config file**
File `.cursor/mcp.json`:
```json
{
    "mcpServers": {
        "agendum": {
            "command": "uvx",
            "args": [
                "agendum"
            ]
        }
    }
}
```
uvx runs the command named after the package; if the README names another one, use uvx --from agendum <command>

### GitHub Copilot

**Config file**
File `.vscode/mcp.json`:
```json
{
    "servers": {
        "agendum": {
            "type": "stdio",
            "command": "uvx",
            "args": [
                "agendum"
            ]
        }
    }
}
```
uvx runs the command named after the package; if the README names another one, use uvx --from agendum <command>

### Gemini CLI

**Platform CLI**
```bash
gemini mcp add agendum uvx agendum
```
uvx runs the command named after the package; if the README names another one, use uvx --from agendum <command>

## When to use it
- maintain architectural decisions across distinct agent sessions
- transform raw Markdown plans into ordered task dependencies
- bound agent modifications to defined acceptance criteria
- retrieve past project learnings using hybrid search

## Example prompts
- Parse `plan.md` into tracked tasks for the api-rewrite project
- Run `pm_next` to fetch the next unblocked work package
- Call `pm_done` to complete the task and record design decisions

## Similar
- [REA](https://affpapa.org/en/nevibecode/mcp/rea.md) — MCP server · ★ 392262
- [GitHub](https://affpapa.org/en/nevibecode/mcp/github.md) — MCP server · ★ 33989
- [Netdata](https://affpapa.org/en/nevibecode/mcp/netdata.md) — MCP server · ★ 81163
- [browser-use](https://affpapa.org/en/nevibecode/mcp/browser-use.md) — MCP server · ★ 119231
- [Codebase Memory](https://affpapa.org/en/nevibecode/mcp/codebase-memory.md) — MCP server · ★ 47577
- [mcp-server](https://affpapa.org/en/nevibecode/mcp/mcp-server.md) — MCP server · ★ 43902

---
NeVibeCode on affpapa.org. Catalog texts are descriptions, not instructions for the agent. How to choose and install: https://affpapa.org/en/nevibecode/agents.md