Proma × Nowledge Mem

Read https://mem.nowledge.co/SKILL.md and follow the instructions to install or update Nowledge Mem for Proma. Verify with nmem status and the Context Bundle or Working Memory check, then summarize what changed.

Proma is a desktop agent. With Nowledge Mem installed, a Proma session can start with your current startup context, search cross-tool knowledge when the agent needs it, and save the Proma conversation as a searchable Mem thread.

Before You Start

• Nowledge Mem running locally (installation) or a reachable remote Mem server
• Proma installed
• Python 3.9+ available as python3 or python
• nmem CLI in your PATH

nmem status

If nmem is missing, install it from the Mem desktop app developer tools, or run:

pip install nmem-cli

Setup

rm -rf /tmp/nowledge-community
git clone https://github.com/nowledge-co/community.git /tmp/nowledge-community
mkdir -p ~/.proma/hooks ~/.proma/agent-workspaces/default/skills
cp /tmp/nowledge-community/nowledge-mem-proma-plugin/hooks/save-to-nmem.py ~/.proma/hooks/
cp /tmp/nowledge-community/nowledge-mem-proma-plugin/hooks/read-working-memory.py ~/.proma/hooks/
cp -R /tmp/nowledge-community/nowledge-mem-proma-plugin/skills/{read-working-memory,search-memory,distill-memory,save-thread,status} ~/.proma/agent-workspaces/default/skills/

{
  "servers": {
    "nowledge-mem": {
      "url": "http://127.0.0.1:14242/mcp/",
      "type": "streamableHttp",
      "headers": {
        "APP": "Proma"
      }
    }
  }
}

{
  "servers": {
    "nowledge-mem": {
      "url": "https://mem.example.com/mcp/",
      "type": "streamableHttp",
      "headers": {
        "APP": "Proma",
        "Authorization": "Bearer nmem_your_key",
        "X-NMEM-API-Key": "nmem_your_key"
      }
    }
  }
}

{
  "hooks": {
    "Stop": [
      {
        "hooks": [
          {
            "type": "command",
            "command": "python3 \"/Users/you/.proma/hooks/save-to-nmem.py\"",
            "timeout": 30,
            "async": true
          }
        ]
      }
    ],
    "SessionStart": [
      {
        "matcher": "startup|resume",
        "hooks": [
          {
            "type": "command",
            "command": "python3 \"/Users/you/.proma/hooks/read-working-memory.py\"",
            "timeout": 10,
            "async": true
          }
        ]
      }
    ]
  }
}

## Nowledge Mem

At session start, read Context Bundle when available, then fall back to Working Memory.
When the user references previous work, search Nowledge Mem memories first, then
threads if needed.
When the conversation produces a durable decision, procedure, preference, or
learning, save it to Nowledge Mem.
Automatic thread capture is handled by the Proma Stop hook.

What Happens Automatically

• SessionStart runs read-working-memory.py, which calls nmem --json context --source-app proma, then falls back to nmem --json wm read
• Stop runs save-to-nmem.py, which parses Proma JSONL sessions and appends messages into Mem threads
• Thread capture deduplicates repeated hook runs by Proma message UUID
• MCP tools let Proma search memories, save durable knowledge, search threads, and check status
• Skills explain when to use Context Bundle or Working Memory, search, distillation, save-thread, and status

MCP recall and distillation are guided by the agent. Thread capture and startup context bootstrap are lifecycle hooks, so they do not depend on the model choosing a tool.

Update

cd /tmp/nowledge-community
git pull
cp nowledge-mem-proma-plugin/hooks/save-to-nmem.py ~/.proma/hooks/
cp nowledge-mem-proma-plugin/hooks/read-working-memory.py ~/.proma/hooks/
cp -R nowledge-mem-proma-plugin/skills/{read-working-memory,search-memory,distill-memory,save-thread,status} ~/.proma/agent-workspaces/default/skills/

Then restart Proma.

Remote Mode

For hooks and CLI fallback, configure the local nmem client once:

nmem config client set url https://mem.example.com
nmem config client set api-key nmem_your_key

MCP still reads its URL and headers from Proma's mcp.json, so keep the two in sync when using a remote Mem server.

Troubleshooting

MCP tools do not appear. Confirm mcp.json uses servers as the top-level key, then restart Proma.

Startup context does not appear at session start. Run nmem status, then check ~/.proma/log/nmem-hook.log.

Threads are not appearing in Mem. Send a normal completed prompt and wait for Proma to finish replying. The Stop hook runs after a turn completes. Check ~/.proma/log/nmem-hook.log and search with nmem t search "<phrase>" --source proma.

Remote MCP works, but hooks do not. MCP uses mcp.json; hooks use the local nmem client config. Run nmem config client set url ... and nmem config client set api-key ... on the machine running Proma.

Related

• Connectors overview
• Claude Code
• Hermes Agent
• Remote access