chore: merge origin/main into mintlify/af82bda4 and resolve conflicts
This commit is contained in:
@@ -0,0 +1,26 @@
|
||||
---
|
||||
name: forget
|
||||
description: Delete the Mem0 memories stored for this repository and this user. Use when the user asks to forget, clear, wipe, or delete memories.
|
||||
disable-model-invocation: true
|
||||
---
|
||||
|
||||
# Forget this repository's memories
|
||||
|
||||
This permanently deletes remote memories. Before running anything, tell the
|
||||
user exactly what will be deleted: their own memories for this repository
|
||||
only. The repository's project memory is shared by everyone who works in it,
|
||||
so it stays unless the user explicitly asks to delete that too.
|
||||
|
||||
After the user confirms, run:
|
||||
|
||||
```bash
|
||||
python3 "${PLUGIN_ROOT}/core/memory_cli.py" --harness "codex" --plugin-data-dir "${PLUGIN_DATA}" forget --remote --yes
|
||||
```
|
||||
|
||||
If the user also asked to delete the repository's shared project memory, add
|
||||
`--include-project-memory` and say that this removes it for every teammate.
|
||||
|
||||
Report what the command output says was deleted. If the user only wants local
|
||||
data cleared (evidence log, pending queue), run the same command without
|
||||
`--remote`. Never pass `--yes` before the user has confirmed in this
|
||||
conversation.
|
||||
@@ -0,0 +1,20 @@
|
||||
---
|
||||
name: pause
|
||||
description: Pause Mem0 memory capture on this machine. Use when the user wants to stop memories being recorded, for example for private work or experiments.
|
||||
disable-model-invocation: true
|
||||
---
|
||||
|
||||
# Pause memory capture
|
||||
|
||||
To pause (hooks stop capturing and sending session content; a minimal
|
||||
anonymous telemetry ping still fires at session start unless
|
||||
`MEM0_TELEMETRY=false`):
|
||||
|
||||
```bash
|
||||
python3 "${PLUGIN_ROOT}/core/memory_cli.py" --harness "codex" --plugin-data-dir "${PLUGIN_DATA}" pause
|
||||
```
|
||||
|
||||
Confirm the new state back to the user, and remind them that already-created
|
||||
memories still exist and remain searchable. Pending unsent packets are held
|
||||
while paused, not expired, and are delivered after resuming. To turn capture
|
||||
back on, use `/mem0:resume`.
|
||||
@@ -0,0 +1,21 @@
|
||||
---
|
||||
name: remember
|
||||
description: Acknowledge a "remember this" request and make sure it is captured well. Use when the user explicitly asks to remember, note, or save something for future sessions.
|
||||
disable-model-invocation: true
|
||||
---
|
||||
|
||||
# Remember something for future sessions
|
||||
|
||||
Mem0 creates memories from the session automatically — there is no separate
|
||||
write command. When the user asks to remember something:
|
||||
|
||||
1. Restate the fact clearly and completely in your reply, in one or two
|
||||
sentences, including any names, values, or paths it depends on. Your visible
|
||||
reply is what memory extraction reads, so a precise restatement is what gets
|
||||
remembered.
|
||||
2. Tell the user it will be saved with this session's memories when the session
|
||||
ends or compacts, and that it will surface in future sessions in this
|
||||
repository (they can check later with /mem0:search).
|
||||
|
||||
Do not invent a storage confirmation or a memory ID — creation happens in the
|
||||
background after the session.
|
||||
@@ -0,0 +1,19 @@
|
||||
---
|
||||
name: resume
|
||||
description: Resume Mem0 memory capture after it was paused with /mem0:pause.
|
||||
disable-model-invocation: true
|
||||
---
|
||||
|
||||
# Resume memory capture
|
||||
|
||||
Resume memory capture for this machine.
|
||||
|
||||
Run:
|
||||
|
||||
```bash
|
||||
python3 "${PLUGIN_ROOT}/core/memory_cli.py" --harness "codex" --plugin-data-dir "${PLUGIN_DATA}" resume
|
||||
```
|
||||
|
||||
Confirm to the user that capture is active again. New sessions record evidence and
|
||||
create memories as normal; nothing that happened while paused is retroactively
|
||||
captured.
|
||||
@@ -0,0 +1,28 @@
|
||||
---
|
||||
name: search
|
||||
description: Search memories from earlier Codex sessions in this repository. Use it when earlier work may already explain the code, error, decision, or command you need, so you can avoid repeating file reads, searches, or experiments.
|
||||
argument-hint: "[question] [--top-k number] [--category category-name] [--scope repo|dir|mine] [--run-id session-id]"
|
||||
disable-model-invocation: true
|
||||
---
|
||||
|
||||
# Search memories
|
||||
|
||||
Call `search_memories` with the user's question. Treat `--top-k`, `--category`,
|
||||
`--scope`, and `--run-id` as tool arguments instead of including them in the
|
||||
query.
|
||||
|
||||
Omit `top_k` to use Mem0's configured default. Omit `category` to search every
|
||||
category; a category is a best-effort label Mem0 assigned when it saved the
|
||||
memory, so if a category search misses, repeat it without the category. Omit
|
||||
`scope` to use the configured default, normally `repo`: this repository's
|
||||
shared memory, which everyone who works in it contributes to, plus your own
|
||||
preferences.
|
||||
|
||||
Pass `scope` when the question needs something else: `dir` to narrow the
|
||||
shared memory to the directory you are working in (a package inside a
|
||||
monorepo), `mine` for your own preferences alone.
|
||||
|
||||
Pass `run_id` with any scope to retrieve memories saved in a specific coding-agent
|
||||
session. Omit `run_id` to search across sessions. It filters the memories returned;
|
||||
it does not identify the session making the search request. Use a known session ID,
|
||||
never invent one. Return the tool's result directly.
|
||||
@@ -0,0 +1,23 @@
|
||||
---
|
||||
name: status
|
||||
description: Show whether Mem0 memory is working in this repository, covering configuration, capture state, pending flushes, and whether the Mem0 API key is valid. Use when the user asks whether memory is on, why a memory is missing, or anything looks broken.
|
||||
disable-model-invocation: false
|
||||
---
|
||||
|
||||
# Memory status
|
||||
|
||||
Run both commands and report the combined result in plain language:
|
||||
|
||||
```bash
|
||||
python3 "${PLUGIN_ROOT}/core/memory_cli.py" --harness "codex" --plugin-data-dir "${PLUGIN_DATA}" status --json
|
||||
python3 "${PLUGIN_ROOT}/core/memory_cli.py" --harness "codex" --plugin-data-dir "${PLUGIN_DATA}" doctor
|
||||
```
|
||||
|
||||
Summarize, using only fields the JSON actually reports: whether capture is
|
||||
active or paused, the user ID and repository scope (`repo_id`), whether an
|
||||
API key is configured, the event/flush/retrieval counts (`flushes` is the
|
||||
number of completed flushes, not a pending count), and the doctor check
|
||||
results. If doctor reports an authentication failure (401 / invalid key), say
|
||||
clearly that the Mem0 API key is invalid or expired and that memories are NOT
|
||||
being created. Never report an auth failure as "no memories found". Suggest
|
||||
reinstalling with `--config api_key=...` in that case.
|
||||
Reference in New Issue
Block a user