numbat
v0.3.0 open-sourceThe section your link pointed to isn't on numbat's page any more, so we've brought you to the top.
numbat provides endpoint visibility into AI agent activity, with local detection, optional pre-action blocking, and forensic reconstruction.
numbat scan --agent muse-code numbat agents --all numbat hook install --agent codex --emit all numbat rules check --rules-dir ./numbat-policy && numbat hook install --agent codex --emit all --rules-dir ./numbat-policy --enforce
Summary
numbat is an open-source, single-binary command-line tool that watches AI agents running on a laptop or server and reconstructs what they did. It's free to use, distributed as a binary for macOS, Linux, and Windows, or installable via go install, with no server component required. It works through local hooks and plugins for live agents like Codex, OTLPOTLPOpenTelemetry Protocol, a standard way to send traces, metrics and logs between applications, collectors and monitoring systems. It lets telemetry move across compatible tools./HTTP log ingestion, and read-only scanning of on-disk session artifacts, normalizing everything into one event model that a local CELCELCommon Expression Language, an open-source expression language maintained by Google, used to embed lightweight, safe policy evaluation logic into tools without running a full scripting engine. rule engine evaluates for detection and, optionally, pre-action blocking. It's built for teams securing AI agent deployments — platform and detection engineers rather than end users — who need endpoint visibility into agent behavior rather than network- or cloud-layer monitoring. Findings and enforcement decisions are written as versioned NDJSONNDJSONA text format where each line is a self-contained JSON object, making it easy to stream, append, and parse large datasets record-by-record without loading the whole file into memory., with case bundles for investigation.
numbat provides endpoint visibility into AI agent activity, with local detection, optional pre-action blocking, and forensic reconstruction.
How numbat helps
Which AI agents are actually running on this machine right now?
reports discovered agents by name with their config signals, on-disk artifacts, live capture mode, and whether a hook is already wired, before anything is installed
What did an agent actually do during a past session?
reconstructs a chronological per-session timeline from on-disk session artifacts alone, with no prior instrumentation required
Did an agent try something that matches a known risk pattern?
evaluates normalized events against local CEL rules, including multi-step sequence rules, and writes findings with the source event references that triggered them
Would a custom policy have actually caught the behavior I'm worried about?
validates and test-runs operator-authored YAML rules against a saved NDJSON fixture, showing what the rule would have flagged before it's ever loaded live
Can I stop an agent from writing to a sensitive file before it happens?
supports opt-in pre-action blocking on supported synchronous hooks, but only for rules explicitly marked to enforce — every shipped rule stays monitor-only until an operator opts it in
Can I hand this incident to someone else without exposing the whole transcript?
builds a portable case bundle with a SHA-256 manifest whose records redact secrets and omit raw transcripts by default, keeping raw evidence out unless explicitly added
Languages
Command line
No command matches.
| option | found in | since | description |
|---|
Values are placeholders taken from each option’s declared default. Nothing is executed here — the output shown is a recording of a run that already happened.
Score card
-
Shipping velocity releases a month, last 12 months
-
Documentation quality release-notes rubric, 0–100
-
Breaking-change discipline how rarely a release breaks users, and notice given
-
Maintenance risk quiet time against its own cadence, maintainers
-
Innovation & surface new capabilities, API and MCP surface
- Community health
-
Contributor trend distinct committers, this year against last
-
Maintainer concentration bus factor now against a year ago
-
Issue backlog trend issues closed per issue opened, 12 months
-
Responsiveness median time to a maintainer’s first reply
Nine scores for numbat, each built from the releases, notes and API history this site tracks — and, on Researcher and Business, where it stands among every measured tool in its category and how that has moved.
Score cards are part of every paid plan.
See plans → Already subscribed? Sign in →Trajectory
Blank means never measured; nothing here is a filled-in zero.
Peer standings are included with Researcher and Business. See how this tool ranks against its competitors, using data we don't publish anywhere else.
See plans →Each score is 0–100 on a published rubric, derived from data this site already shows. A dimension we cannot measure says so — it is never scored 0 and never enters a rank.
How this is scored
- Shipping velocity — releases a month over the last 12 months, on a log scale where 16 a month scores 100. Needs 3 releases in the window.
- Documentation quality — the release-notes rubric published on the analytics page, unchanged.
- Breaking-change discipline — the share of releases carrying a breaking change (none scores 100, half or more scores 0). Only projects that label breaking changes are measured: a clean record cannot be told from an unlabelled one.
- Maintenance risk — quiet time measured against the tool's own release cadence (quiet longer than 90 days and three typical gaps scores 0), plus the number of active maintainers and whether contributors are growing, where known. Higher is lower risk.
- Innovation & surface — new capabilities in the feature ledger this year, API and MCP operations added, and the share of commits that are features: the mean of whichever we hold.
- Community health (open-source tools only) — each is a direction, not a size: contributors this year against last, the bus factor now against a year ago, and issues closed per issue opened, where steady scores 50, doubled 100 and halved 0; plus the median time to a maintainer's first reply (an hour scores about 100, a month or more 0). A commercial card has no repository and carries none of the four.
- Documented delivery (commercial tools) — when a vendor's documented new capabilities have slowed against its own baseline, that counts against the overall standing as heavily as two low dimensions. Steady delivery, or too little history to judge, leaves the standing alone; it is never a sixth score.
- Open-source tools are scored on repository signals; commercial tools on their published changelog. Ranks compare tools within one category and one profile and need at least five measured peers; the overall standing needs three ranked dimensions. Best-in-breed or consolidator is a shape, not a quality, and is never scored.
Release-note detail over time
See how much detail numbat puts in its release notes, quarter by quarter, and compare it with tracked tools in the same category. This history uses up to ten retained, reviewed notes per tool; sparse or blank older quarters may reflect that limit.
Checking release-note history…
Coverage
Missing something? Tell us
What numbat connects to, runs on and detects, read from its release notes and docs. So far: 1 new integration.
1 change, tracked since Oct 19, 2026.
-
Oct 192026
new Muse Code live-hook and at-rest session support
**Muse Code support:** Live monitoring, opt-in enforcement, and forensic scanning of saved sessions (#59).
Editors & IDEs 1
Nothing first seen in this window. That doesn't mean numbat shipped nothing, only that we found no record of it.
How we count
Read from 1 of 3 release notes and docs updates we hold for numbat. "First seen" is the earliest note that names something; "last" is the latest. Our history may start partway through numbat's life, so an old integration can show a recent first-seen date.
Each change counts once: "added 40 checks" is one line, so the numbers measure how often coverage moves, not how big the catalogue is or how well it detects.
Releases
3 entries
No entries match these filters.
- October 20261
- August 20261
- July 20261
Project health
Issue, contributor and maintainer activity in perplexityai/numbat, 2023-10 – 2026-09, as of 2026-09-25. Twelve-month figures first; the trend lines run the whole window.
- Contributors 2 +2 vs prior 12 months
- Maintainers 1 +1 joined · −0 left, last 12 months
- Bus factor 1 committers behind half the commits
- Community-filed 100% of issues opened, last 12 months
- Median time to close 11 days first maintainer reply 5.4 days
- Open backlog 6 of 11 issues ever filed
- Issues opened / month 2 · 0 communitymaintainers
- Open backlog 6
- Contributors / month 0
- Commits / month 0
- Stars 1,062
- other 100%
Counts from the project's public repository and git history; no individuals are identified. Maintainers are the committers covering 80% of the last year's commits. how we measure this
Add the badge to your README
Maintain this project? The badge links readers straight to this page, so they can follow every release here.