agentsclimarketplace

En US

Skill Zoranner/memo-brain/skills/memo-brain/en-US

Manage and retrieve cross-conversation memory. Public action semantics follow command-philosophy. Use for "remember this", "search memory", "organize memory", or "show current state".From its SKILL.md

Install
npx -y skills add Zoranner/memo-brain --skill en-US

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

  • no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
  • 3 stars3 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

5.2 KB, ~1.2k tokens by cl100k_base, as published. Nobody here has run it

Memo Brain Management

This skill follows the public action language defined in docs/architecture/command-philosophy.md.

Standard Actions

  • memo awaken
  • memo remember
  • memo recall
  • memo reflect
  • memo dream
  • memo state

Current Capability Boundaries

The current CLI does not provide these old interfaces. Do not reason or act as if they still exist:

  • memo embed
  • memo search
  • memo restore
  • memo update
  • memo merge
  • memo delete
  • memo list
  • --tags
  • --after / --before

If the user speaks in that old product language, translate it into the standard action semantics. If the current system cannot support the request, say so directly instead of fabricating capabilities. search / embed may appear only as natural-language triggers or warnings about old terminology, not as runnable command examples.

When to Use

Use this skill when:

  • The user explicitly asks to remember or record something
  • The user wants to search or recall past memory
  • You need details for one memory record
  • You need to run dream / maintenance
  • You need current engine state
  • You need dream-driven derived-layer maintenance

Do not use this skill when:

  • The task is ordinary code search inside the repository
  • The task does not need cross-conversation memory
  • The request depends on update/merge/delete/list behaviors that do not exist in the current CLI

Recommended Workflow

Awaken a Memory Space

memo awaken

Remember Content

Standard action: remember

memo remember "<content>"

If you already know structured information, add it explicitly:

memo remember "<content>" --entity person:Alice --entity place:Paris --fact Alice:lives_in:Paris

Recall Content

Standard action: recall

memo recall "<query>" -n 10

If fast-path retrieval is likely insufficient:

memo recall "<query>" -n 10 --deep

Reflect on One Memory

Standard action: reflect

memo reflect <memory-id>

Dream

Standard action: dream

memo dream

For full derived-layer maintenance:

memo dream --full

Inspect State

Standard action: state

memo state

How to Choose the Action

User IntentStandard ActionCurrent Execution
"remember this conclusion"remembermemo remember ...
"did we solve something like this before"recallmemo recall ...
"show me that memory in detail"reflectmemo reflect ...
"organize the memory"dreammemo dream
"what is the current system state"statememo state
"indexes may be inconsistent, restore them"dreammemo dream --full

Retrieval and Recording Principles

Recording Principles

  • Record durable experience, facts, decisions, or troubleshooting outcomes worth keeping
  • Focus on the content first; add --entity and --fact when you can do so concretely
  • Do not design workflows around nonexistent features such as tags, update, merge, or list
  • Default remember does not call providers; do not make embedding writes the default recording premise

Retrieval Principles

  • Queries should include situation and intent, not only loose keywords
  • Start with default memo recall
  • Only use --deep when default results look weak, the topic spans multiple layers, or the user explicitly asks for deeper recall
  • Use memo reflect when you need detail on one returned record
  • Default recall does not call providers; do not claim full semantic retrieval when providers are absent

Maintenance Principles

  • Derived-layer maintenance goes through memo dream; use memo dream --full only when full maintenance is needed
  • dream may enter extraction / embedding slow paths; do not present those slow paths as default remember / recall behavior
  • Use Working Set / Pinned for user-facing memory semantics; do not expose L0/session as the user's mental model

Common Mistakes

Don'tDo
Keep calling memo search / memo embedTranslate into standard action semantics
Pretend old commands are still the standardUse awaken/remember/recall/reflect/dream/state directly
Fake update/merge/delete/list capabilitiesState directly that they are not implemented in the current CLI
Treat extract as the main memory entrypointOrganize the workflow around public actions only
Treat memo restore as standard maintenanceUse memo dream, or memo dream --full when needed
Explain user-visible state with L0/sessionUse Working Set / Pinned

Trigger Phrases

ActionTrigger Phrases
remember"remember this", "record this", "save this experience"
recall"how did we do it before", "search memory", "do you remember"
reflect"show that memory in detail", "open the details"
dream"organize memory", "run dream", "restore derived layers", "restore index state"
state"show current state"

For executable examples, see examples.md.

What ships with it: 1 file

2.3 KB alongside SKILL.md

Keep looking

Skills are one crate of 326,851. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.