---
name: Nowledge Mem
slug: nowledge-mem-3
category: AI Engineering
description: Nowledge Mem helps you resume work with startup context, search past threads and memories, and save durable knowledge. Use it when you need exact prior decisions, preferences, or CodeBuddy transcript imports.
github: "https://github.com/nowledge-co/community/tree/main/nowledge-mem-codebuddy-plugin/skills/nowledge-mem"
language: Python
stars: 166
forks: 35
install: "npx degit https://github.com/nowledge-co/community/tree/main/nowledge-mem-codebuddy-plugin/skills/nowledge-mem ~/.claude/skills/nowledge-mem"
installs_to: ~/.claude/skills/nowledge-mem
source_path: nowledge-mem-codebuddy-plugin/skills/nowledge-mem/SKILL.md
collection_size: 25
category_size: 3670
collection_url: "https://dirskills.com/collections/nowledge-co/community"
added: 2026-09-08T05:34:52.715Z
last_synced: 2026-09-08T05:34:52.715Z
canonical_url: "https://dirskills.com/skills/nowledge-mem-3"
---

# Nowledge Mem

Nowledge Mem helps you resume work with startup context, search past threads and memories, and save durable knowledge. Use it when you need exact prior decisions, preferences, or CodeBuddy transcript imports.

**Install:**

```bash
npx degit https://github.com/nowledge-co/community/tree/main/nowledge-mem-codebuddy-plugin/skills/nowledge-mem ~/.claude/skills/nowledge-mem
```

## README

Nowledge Mem is the user's cross-tool memory. Use it to resume with current context, recall exact prior work, save durable knowledge, and make CodeBuddy sessions searchable elsewhere.

## Startup Context

Read Context Bundle at the beginning of meaningful work or when resuming:

```bash
nmem --json context --source-app codebuddy
```

If an older CLI rejects `context`, use `nmem --json wm read`. Do not read both unless the user asks.

## Recall

Search when the user references prior work, resumes a project, investigates a regression, asks for rationale, or needs exact conversation history.

Prefer MCP:

- `memory_search` for decisions, preferences, procedures, and learnings.
- `thread_search` for prior conversations.
- `thread_fetch_messages` only after identifying a relevant thread.

CLI fallback:

```bash
nmem --json m search "what to recall"
nmem --json t search "conversation to find" --source codebuddy -n 5
```

## Save Durable Knowledge

Search before saving. Update an existing memory when the new information evolves it; add a new memory only for a durable decision, reusable procedure, stable preference, correction, or non-obvious learning.

```bash
nmem --json m search "existing concept"
nmem --json m add "content" -t "Title" --unit-type decision -l "label" -s codebuddy -i 0.8
```

## CodeBuddy Threads

Lifecycle hooks automatically sync the active transcript. For a deliberate historical import:

```bash
nmem t sync --from codebuddy --limit 20
nmem t sync --from codebuddy --apply
```

Real capture runs on the CodeBuddy machine because MCP cannot read local files under `$CODEBUDDY_CONFIG_DIR/projects` or `~/.codebuddy/projects`.

## Status

```bash
nmem --json status
```

If status succeeds but CodeBuddy-specific commands fail, update the same CLI installation before debugging hooks or MCP.

Respect `NMEM_AGENT_ID`, `NMEM_HOST_AGENT_ID`, and `NMEM_SPACE` when present. `source_app=codebuddy` is provenance, not an AI identity.