Memorb
Agent-Agnostic Second Brain memory framework inspired by movie and library science.
npx -y skills add iamjosuho/memOrb --skill memorbAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
3 things to look at
- 12 days oldThe repository was created 12 days ago. New is not bad, but a brand new repository carrying a familiar-sounding name is the shape a typosquat arrives in, and there has been no time for anyone else to find a problem with it.
- 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.
- 1 stars1 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.
What its author says it does
Copied from the file, not written here
Gateway skill for Second Brain operations (read/write orbs, OrbTrack capture & triage, ingest, query, lint, archive, memory replay, meetings, business cards). Read first to route to appropriate sub-skills.
SKILL.md
4.9 KB, as published. Nobody here has run it
Memorb (Gateway Skill)
CRITICAL: Primary routing hub for the vault. Every file operation must strictly follow these instructions and sub-skill routing without exception.
Instruction Priority
- Explicit User Instructions (highest)
- Memorb & Routed Sub-skills (overrides default agent behavior)
- Default Agent Behavior (lowest)
Core Rules
Before performing any file operation:
- Read this gateway file.
- Read
SKILL.mdfor every matching sub-skill in the routing table below. - All write operations must adhere to
memorb-conventions(paths, naming, frontmatter). - Load multiple sub-skills if needed (e.g., transcript processing =
m365-meeting-note+memorb-ingest).
Ownership boundary
memOrb may read anything the user points it at; it writes only inside memorbs/. No exceptions — this binds extensions too.
Reading is safe: no state change, no surprise. Conditional writing is where the trouble lives. The moment a skill says "write to the reference folder if it exists", the same operation produces three different outcomes depending on a coincidence of folder naming — it writes for one user, silently skips for the user who numbered their folders, and does nothing for the user who has none. That is worse than either owning the folder or never touching it.
So memOrb grows exactly one folder in the user's vault: memorbs/. Delete it and the vault is clean.
Skill Router Table
Core skills reside in skills/core/{name}/SKILL.md.
Extensions reside in skills/extensions/{name}/SKILL.md.
Core Skills (memorb-core)
| Sub-skill | Location | Trigger / Scenario |
|---|---|---|
| memorb-conventions | core/memorb-conventions | Base layer: paths, directory structure (memorbs/), naming, frontmatter, templates. Required for ALL write ops. |
| memorb-born | core/memorb-born | Vault initialization (Phase 0), persona seed setup, CLAUDE.md generation, or /memorb-born reset. |
| orbtrack-triage | core/orbtrack-triage | Process OrbTrack (memorbs/HQ/OrbTrack/), PARA organization, note archiving |
| island-reclamation | core/island-reclamation | Create new Island (long-term interest/responsibility domain): discuss structure before file creation |
| dream-studio | core/dream-studio | Monthly memory replay: propose Core orbs, add/revise Belief orbs, refresh Island narrative. Sole writer of memorbs/HQ/Core/. |
| memorb-ingest | core/memorb-ingest | Ingest raw materials (transcripts, articles, PDFs) into memorbs/Long-Term//memorbs/ wiki |
| memorb-query | core/memorb-query | Q&A, comparisons, decision recommendations via memorbs/Long-Term//memorbs/ wiki |
| memorb-lint | core/memorb-lint | Vault health check and memorbs/Long-Term//memorbs/ wiki MUSTY linting |
| memorb-forgetter | core/memorb-forgetter | Execute MUSTY archiving to memorbs/Dump/ and update vault Wiki Links |
| writing-memorb-skills | core/writing-memorb-skills | Create/modify vault skills (Meta) |
Extension Skills (memorb-extensions)
| Sub-skill | Location | Trigger / Scenario |
|---|---|---|
| m365-meeting-note | extensions/m365-meeting-note | Process Teams meeting transcripts & minutes |
| recording-transcription | extensions/recording-transcription | Non-Teams voice recordings (phone, interview, m4a) to transcript |
| business-card-ingestion | extensions/business-card-ingestion | Business cards to memorbs/Long-Term/People/ Page Bundle |
| memorb-domain-query | extensions/memorb-domain-query | Bi-directional domain/email lookup & M365/Outlook integration |
Extension Interactive Installation & Fallback Protocol
If a scenario calls for an extension (e.g. business card OCR, M365 meeting transcript processing), check if skills/extensions/{name}/SKILL.md exists:
- Installed: Read the extension's
SKILL.mdand execute its specialized workflow. - Not Installed (Interactive Installation):
- Ask the user:
"💡 Notice: The '{extension-name}' extension provides enhanced domain formatting for this task. Would you like me to install it for you?" - User Approves: Install the extension into
skills/extensions/{extension-name}/and execute its specialized workflow immediately. - User Declines / Skip: Proceed with the graceful fallback to core
memorb-ingestormemorb-query.
- Ask the user:
Red Flags
Stop and revert if rationalizing any of the following:
- "It's simple, no need to read sub-skills" → Router rules are mandatory for all file ops.
- "Place notes in root temporarily" → New notes belong in
memorbs/HQ/OrbTrack/or PARA folders. - "I recall the rules without reading" → Always re-read SKILL.md as rules evolve.