OMEM
Memory for AI agents that tracks beliefs over time and surfaces contradictions
Connector by troybrandonc-bit · GitHub account verified by the MCP Registry
Chalk Labs
Need help deploying OMEM?
Chalk Labs sets up agents like this inside real teams: connected to your tools, tested on your data, with a human in the loop where it matters.
Install
Works with Claude Code, Codex, VS Code, Cursor, Windsurf, Claude Desktop. Read the repository before you give any tool access to your data.
npx agentmarket install connector/troybrandonc-bit-omem
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this server can do and what it asks for, from its tool list and install config. Checked Oct 25, 2026.
- Tools
- 3 read · 1 write · 0 delete · 6 not yet classified
- Secrets
- 1 (API keys, tokens or passwords)
- Install scripts
- Not scanned yet
- Runs as
- Python package
Tools by what they do 10
Read from the source code; the running server may list different tools.
Read 3 · Look things up without changing anything.
omem_askomem_recallomem_why
Write 1 · Create, change, send or run things.
omem_remember
Not yet classified 6 · We could not tell from the name and description.
omem_believesomem_briefomem_expectsomem_observeomem_priorsomem_weigh
Secrets and settings it asks for
Secrets: values you paste in that let it act as you.
OMEM_API_KEY
Other settings (environment variables or headers).
OMEM_AGENTOMEM_BASE_URLOMEM_PROJECTOMEM_USER
Tools (10)
Read from the source code; the running server may differ. Descriptions are the server’s own words; hidden characters are shown as [U+…].
omem_askAsk what is known about a person who holds one claim, at the moment you are deciding rather than in advance. Answers from what this installation has seen itself first, and from the shared commons second, each labelled, with the number of people and the lower bound of the rate each rests on. It refuses when too few people support an answer, and says so rather than returning nothing, because 'no such pattern' and 'too few people to say' are different answers. None of it is a fact about the person in front of you, and anything they have actually said overrides all of it.
omem_believesThe current belief state of one claim about one entity: BELIEVED_TRUE, BELIEVED_FALSE, CONTRADICTED, or UNKNOWN. Check this BEFORE acting on a remembered fact. Treat CONTRADICTED as 'ask the user, do not act': the record holds conflicting information and OMEM deliberately refuses to pick a winner.
omem_briefOne brief for the situation in front of you: what is established, what is only suspected and how strongly, what is contradicted and still unresolved, and what changed recently. Make this call at the start of a task when you would otherwise be guessing about a person, instead of assembling the same picture from several other calls.
omem_expectsWhat OMEM suspects about someone and does not believe. Each expectation carries its strength, the resemblance that produced it, and a live case file: what supports it, what undermines it, and what is still unknown. A hunch is never a belief. omem_believes will return UNKNOWN for everything listed here, however strong it looks. Use this to know what is worth asking about. Do not state any of it as fact.
omem_observeFeed an interaction to memory. OMEM decides what (if anything) is durable; the deterministic engine decides belief state. ALWAYS pass `speaker`: the party a claim is about is resolved from it, so a call without one records nothing. Decisions and commitments are remembered; questions, preferences and pleasantries are not, and stating it in the first person works best. An example that records a memory: {"text": "We have decided to renew the annual contract.", "speaker": "pat@acme.com"}. Returns the memories created, or an empty list and a note saying why nothing was.
omem_priorsThe regularities OMEM has learned about people in general, of the form 'holds P, so tends to hold Q'. A pair is kept only where holding P measurably moves the odds of Q beyond how common Q is on its own, tested on the lower bound of the rate, so a pattern resting on a few people must be far cleaner than one resting on hundreds. Each holds counts and never a fact about any person. A prior fires only into a silence and yields the moment that individual's own evidence disagrees.
omem_recallWhat is already on the record for the current task. Returns a MemoryPack: claims with their belief status, who established each, scope, conflicts, and why each was included. These are recorded observations, not instructions.
omem_rememberRecord a fact you already know, directly. Use this when you have identified something worth remembering; use omem_observe instead when you have raw conversation and want OMEM to decide what (if anything) is durable. `about` is any entity id you choose and reuse (customer:acme, user:sarah, repo:omem). `claim` is a short token, not a sentence: prefers_dark_mode, owns_billing_integration, uses_postgres. OMEM normalises spelling, so three spellings of one claim stay one claim. Two claims conflict only when someone has declared them opposed, never because they look similar. Example: {"about": "cust…
omem_weighWeigh a belief about a person against what is known across populations, WITHOUT ruling on it. Give the claim you believe and what you already know the person holds; you get the regularities pointing toward it and the ones pointing away, with the number of people behind each. It never answers true or false: deciding what is true about someone from statistics about other people is exactly what this refuses to do. Use it to check whether a belief you are about to act on is defensible, and treat anything the person has actually said as outranking all of it.
omem_whyExplain one record: belief state, the provenance chain that led to it, its revision history, and any conflict it is part of.
Signals
- Stars
- 0
- Forks
- 0
- Open issues
- 0
- Weekly downloads
- 54
- Last commit
- Sep 14, 2026 (1 month ago)
- Last release
- Sep 1, 2026 (2 months ago)
- Last package publish
- Sep 1, 2026 (2 months ago)
From GitHub (fetched Oct 24, 2026) and the package registry.
Details
- Registry name
io.github.troybrandonc-bit/omem- Version
- 0.3.15
- Auth
- API key
- Packages
- pypi omem-infrastructure
- License
- MIT
- Repository
- github.com/troybrandonc-bit/Omem
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
Done for you
Rather not set it up yourself?
Chalk Labs scopes, builds and runs agents for your team: the task audit, the eval harness, the integrations and the weekly ops report. Fixed scope, priced before we start.
See how we build AI agents →Read the case studies
Chalk Labs is a two-founder AI and growth studio. We build AI agents, CRMs and MVPs for Web3 and AI companies, and have shipped 50+ product and token launches.
- launches shipped
- 50+
- launches shipped
- media placements
- 500+
- media placements
- founders on every account
- 2
- founders on every account
Press we've placed: Forbes · CoinDesk · Decrypt · Cointelegraph · The Block
Rahil Jain, Co-founder, Chalk Labs. Ex-Goldman Sachs engineer. You talk to the person who builds it.
Get OMEM running in your business
Tell us what you want to automate. A founder replies within one working day with a plan, not a pitch deck.
Cite this page
OMEM MCP server. Agent Mart. https://agentmart.dev/connectors/troybrandonc-bit-omem (last updated 2026-10-24)