toktop

README

toktop

btop for AI: a terminal dashboard for LLM inference engines and the agents hammering them.

toktop --demo            # simulated fleet, works instantly
toktop                   # auto-discovers local engines (ports + processes)
toktop --agents          # also watch the coding agents on this machine
toktop ssh://maci@box    # watch engines on another host

Install

go install -tags sqlite github.com/maci0/toktop/cmd/toktop@latest

-tags sqlite matches the GitHub binaries and make build: crush and opencode session databases cannot be read without it. Or download a binary for linux, macOS, and Windows (amd64 + arm64) from the releases. An installed binary updates itself in place:

toktop update --check    # report the latest release, change nothing
toktop update            # download it and replace the running binary

AI coding agents

toktop also watches the coding agents running on this machine, not just inference engines. It finds them by process, reads the token counts they already write to their own session logs, and shows their throughput beside the engines:

AGENTS  local, read from their own session logs
  claude   ▲ 1.1k tok/s   ▲2.4k ▼8.1k   ● live
  codex    ▲ 340 tok/s    ▲18k ▼40k     via 127.0.0.1:11434  ● live

--agents turns this on. It is off by default because it means scanning this machine's processes and reading session files nobody pointed toktop at: watching engines you configured does not imply consent to that. The watch needs each process's working directory to attribute transcripts; Windows does not expose that through a documented API, so --agents finds no agents there.

With no engines attached, agents mode is the full dashboard: header rates, throughput charts and the host strip, all driven from the session logs. Prompt, output and (when the agent reports it) reasoning tokens are shown per agent. An agent generating through an engine toktop is already measuring still appears in the list, labelled via <engine>, but those tokens are not added on top of the engine's own numbers.

Once asked for, nothing else has to be configured and the agent does not have to cooperate: claude, codex, qwen, copilot, pi, prime-agent, feynman, clanker, and dsh all keep transcripts that carry the provider's own counts. dsh's default log is concatenated zstd frames; uncompressed JSONL is read too. Agents that report nothing show no rate rather than a zero.

Two agents keep databases instead of transcripts, and both need the sqlite build tag, which decides whether a driver is compiled in at all (released binaries and make build carry it; make build TAGS= leaves it out).

opencode keeps one session store for the whole machine, so it is gated twice: the tag links the driver, and --opencode-db decides whether a binary that has it opens the operator's database. Asking for it in a build without the driver says so on stderr rather than reporting a silent zero.

crush keeps its database inside the project it is working on (.crush/crush.db, at the project root it resolves), with the counts in sessions.completion_tokens. The store is already scoped to that project, so there is no second gate: with the tag, it is read. The only JSONL crush writes is its log, which carries no counters.

Reading is done by this repo's own agentusage package, which gauntlet also imports, so both tools report the same numbers. Agents defined in ~/.gauntlet/agents.json are picked up here too; a malformed file is reported at startup rather than silently shrinking the watch to the built-in agents. Set GAUNTLET_HOME to read that file from somewhere else.

Using the Go package

Import github.com/maci0/toktop/agentusage to discover agent processes and read the token counts they already write:

package main

import (
	"fmt"
	"time"

	"github.com/maci0/toktop/agentusage"
)

func main() {
	if err := agentusage.LoadDefinitions(agentusage.DefinitionsPath()); err != nil {
		fmt.Println(err) // malformed file; a missing one is not an error
	}
	agentusage.EnableOpenCodeDB(true) // no-op without -tags sqlite

	for _, p := range agentusage.Discover() {
		w := agentusage.Watch(p.Tool, p.Dir, time.Now())
		if w == nil {
			continue // this agent keeps no readable transcript
		}
		s := w.Poll()
		if s.Empty() {
			continue
		}
		fmt.Printf("%s pid %d: %d output, %d prompt\n", p.Tool, p.PID, s.Output, s.Input)
	}
}

RegisterSpec teaches the package about an agent it was not compiled to know. errors.Is matches ErrEmptyTool and ErrNoRoots on a rejected spec, and ErrInvalidDefinitions on a malformed definitions file.

What it shows

• Engines - every engine found locally or via ssh, with model, version, KV-cache pressure, queue depth and throughput. Fingerprinted kinds: Ollama, llama.cpp/llamafile/ramalama, vLLM, SGLang, TRT-LLM/Triton, LM Studio, MLX (mlx-lm / LM Studio), KoboldCpp, LocalAI, TGI, LiteLLM, GPUStack, Lemonade, OmniRoute (auto-detected via its routing header; per-model context windows shown) - plus a generic OpenAI-compatible fallback so nothing is left out (text-generation-webui and TabbyAPI are discovered by process/port but identified as generic OpenAI).
• Throughput charts - aggregate decode + prompt tokens/sec as heat-colored area charts; engine-published tok/s gauges are trusted when present. Agents with no local engine (or whose engine is not monitored) add their own rates; tokens already counted by a watched engine are not added again.
• Probes (p, --probe N) - tiny streaming generations measuring real TTFT and decode speed per engine.
• Agent feed - any harness can POST usage events:

  curl -X POST localhost:8420/v1/events -d \
    '{"agent":"coder","kind":"tool","prompt_tokens":4200,"output_tokens":310,"thinking_tokens":40,"note":"shell(git status)"}'
  

  --agents also fills this from local session logs. Per-agent rows show output, prompt and reasoning rates; an agent using a monitored engine is labelled via that engine so its tokens are not added twice.
• System strip - RAM/swap/load, CPU model, OS+kernel, GPU driver versions (incl. CUDA), NPU enumeration (Intel NPU, AMD XDNA NPU, Qualcomm Cloud AI100, Apple Neural Engine with chip generation), GPU temp/util/VRAM/ power - on Apple Silicon including live wired-memory and utilization from IOAccelerator - and a second identity row for sensors.
• Braille charts - dot-matrix rendering with btop-style fading bloom; timescale compresses leftward (t toggles) with faint grid marks showing where each doubling begins.
• Hot reload - on Unix, rebuild the binary while it runs and toktop re-execs into the fresh build (--no-hot-reload to disable). Windows cannot replace a running image: the dashboard exits and asks you to start it again.

Agent feed API

The ingest server runs by default on 127.0.0.1:8420 (--ingest ADDR to move it, --no-ingest to turn it off) and speaks plain HTTP/JSON:

endpoint	purpose
POST /v1/events	record events; body is one JSON object or an NDJSON stream
GET /v1/events	schema hint for humans
GET /healthz	liveness probe, answers ok

Event fields are all optional; anything omitted gets the default:

field	type	default	notes
id	string	-	caller-chosen key, capped at 128 runes; a repeat of a key still in the retained feed (last 64 events) is ignored
ts	RFC 3339 string	arrival instant	offset-less stamps decode as UTC; a space instead of T is accepted
agent	string	anonymous	capped at 64 runes
model	string	-	capped at 128 runes
kind	string	turn	known kinds: turn, tool, error, note; custom kinds pass through lowercased, capped at 24 runes
prompt_tokens / output_tokens / thinking_tokens	integer	0	negative values and values above 2^40 clamp to 0; a whole JSON number such as 100.0 counts; thinking is the reasoning share of output when the agent says so
via_engine	string	-	monitored engine already counting this output; aggregates skip the event; capped at 128 runes
note	string	-	free-form, capped at 512 runes

One POST answers 202 with {"accepted":N} once every event in the stream is recorded, 400 for malformed JSON or a bad ts, 408 when a stream stalls mid-body, and 413 past the 1 MiB body cap. A POST carrying an Origin header (browser-driven; scripts and agents never send one) is refused with 403, so a web page cannot forge rows into a running dashboard. Wrong methods on these paths answer 405 with Allow. Error bodies are short plain-text reasons that name the field or expected shape; unknown fields are ignored, so harnesses can include their own. The request Content-Type header is not checked: the body is always read as JSON/NDJSON, so plain curl -d works unmodified. Every POST is logged to stderr as one structured line (req, status, accepted, duration, remote; failures add error). Event bodies are not logged. Responses carry X-Request-Id, echoed from the request when the sender set one.

Streams are recorded line by line: if a later line fails, events before it stay recorded and the error states how many. Retrying a stream (or a successful POST whose 202 was lost) is safe when each event carries a stable id; without one, replaying the kept lines would duplicate them.

Zero vendor libraries

Everything comes from procfs/sysfs/sysctl, vendor CLIs it shells out to (nvidia-smi, rocm-smi, xpu-smi, system_profiler, ioreg - each the vendor's documented interface with no in-process alternative) or plain HTTP from the engines themselves. SSH transport is an embedded pure-Go client, so remote monitoring needs no ssh binary either. No NVML, no Level Zero, no cgo: single static binary, trivially cross-compiled.

Linux reads /proc + /sys; macOS uses sysctls and system_profiler; Windows uses GlobalMemoryStatusEx, RtlGetVersion and one CIM query for process command lines.

Engines: how discovery works

1. Running processes matching well-known names (ollama, llama-server, vllm, sglang, koboldcpp, lm studio, lemonade-server, ...) give candidate URLs, honouring --port flags.
2. A scan of well-known ports follows (11434, 30000, 8000, 13305, 8080, 1234, 5001, 5000, 4000, 1337, 4891, 7860, ...).
3. Each candidate is fingerprinted by its HTTP surface; anything unrecognized that still speaks OpenAI is shown as such.

Attach anything explicitly:

toktop --add http://10.0.0.5:8000        # repeatable
toktop ssh://user@host                   # remote engines + host vitals

SSH mode is built in (pure Go, no ssh binary needed) and the remote only needs a POSIX shell - no agent is installed. Discovery reads the remote /proc directly: listening sockets from /proc/net/tcp(+6) (with an active port probe as fallback) plus engine processes with their --port flags, so engines on custom ports are found just like locally. Engine traffic rides ssh direct-tcpip channels on that same connection. Each remote engine port is reached through a loopback listener bound to 127.0.0.1 with an ephemeral port, so local clients attach the same way they would to a local engine; those listeners are reachable by any process on this host. Host vitals stream the same way: load, memory, uptime, CPU model, OS, kernel and GPU rows (nvidia-smi, or rocm-smi on AMD boxes).

Auth tries, in order: --ssh-key PATH, keys from ~/.ssh/config (HostName, User, Port, IdentityFile are honored), your default keys, ssh-agent, and finally a password prompt when stdin is a terminal (or set TOKTOP_SSH_PASSWORD for headless runs). Host keys use trust-on-first-use, stored under your config dir; a changed key is refused loudly.

Keys

key	action
q / ctrl+c	quit
esc	close help / quit
space	pause streaming
p	probe every engine with a real generation
t	toggle compressed timescale + grid
? / h	help

Accessibility

toktop is usable without a mouse, without color vision, and with assistive technology:

• Keyboard only - every action has a key (table above); nothing requires pointing or clicking, and ? always shows the full key map.

• Pause freezes everything - space stops the streaming data and the header clock, so a still frame can be read at leisure with a screen reader or magnifier.

• Status never rides on color alone - down engines show ✗ plus their error text, probes show ✓/✗, gauges print their percentage, and the engine count is spelled out numerically in the header.

• Non-visual output - --once prints one static frame instead of running the full-screen UI; a live-repainting dashboard defeats most screen readers, so the static frame is the intended path. Pair it with TOKTOP_COLUMNS / TOKTOP_LINES for a fixed size. --once --plain goes further and prints the same numbers as a linear text report: no braille chart glyphs (which screen readers announce as endless dot-pattern noise or skip entirely), no box-drawing borders, no multi-column panels - just the data in reading order.

  $ toktop --once --plain
  5/5 engines up · out 1.5k tok/s · in 10k tok/s · 2 agents · session 24s
  
  ENGINES
  up   vllm-a100 (vllm)
         Qwen/Qwen2.5-32B-Instruct-AWQ
         out 155 tok/s · in 662 tok/s · kv cache 68% · running 2 · waiting 0 · ttft 115ms
  
  SYSTEM
  memory 64% (248G/384G) · swap 17% · load 5.06
  gpu nv0 A100-SXM4-80GB 82° 69% util vram 57G/80G 397W
  
  PROBES
  ok meta-llama/Llama-3.3-70B-Instruct-engine ttft 113ms 135 tok/s
  
  AGENT FEED
  ops-agent 293 tok/s · research-agent 247 tok/s
  03:29:23 note ops-agent model Qwen/Qwen2.5-32B-Instruct-AWQ prompt 9.2k output 783 note browser(search docs)
  

• Tested contrast - unit tests hold the palette to WCAG 2.2 AA: text colors at >= 4.5:1 on the background, and chart marks at >= 3:1 even at the deepest point of the age fade (internal/ui/theme_test.go).

• No color - NO_COLOR strips styling as usual; layout and text carry the same information without it.

Flags

toktop update     subcommand: install the latest release (--check to only
                  report it, --repo owner/name for a fork)
toktop help       subcommand: usage (help update for the updater's own)
toktop version    subcommand: same as --version
toktop help       same as --help; `toktop help update` for the subcommand
toktop version    same as --version
--demo            simulated fleet, zero setup
--add URL         attach an openai-compatible http(s) endpoint (repeatable;
                  host required, no userinfo; use --bearer / $TOKTOP_BEARER)
ssh://user@host   positional; monitor remote hosts (repeatable)
--ssh-key PATH    private key for ssh targets (overrides ~/.ssh/config;
                  ~ is expanded; a missing file aborts at startup)
--bearer TOKEN    bearer token sent to --add endpoints only; OmniRoute API
                  keys etc. (env: OMNIROUTE_API_KEY, then TOKTOP_BEARER;
                  an explicit --bearer, even empty, wins)
--agents          watch AI coding agents on this machine (session logs)
--opencode-db     with --agents: also read opencode's SQLite session
                  database (needs a build with the sqlite tag)
--probe N         auto-probe every N seconds
--interval D      poll interval (default 1s)
--ingest ADDR     agent event listen address, host:port
                  (default 127.0.0.1:8420; empty is rejected)
--no-ingest       disable the event endpoint (`--ingest` is then ignored)
--once            render one frame and exit
--plain           with --once: linear text report instead of the dashboard
                  frame (screen-reader friendly)
--frames N        with --once: snapshots to accumulate before rendering
--seed N          demo RNG seed
--no-hot-reload   disable restart-on-rebuild while running
--version         print version and exit
--help, -h        show usage, examples and environment fallbacks

Password auth for ssh targets: interactive prompt, or TOKTOP_SSH_PASSWORD.

Environment variables

variable	what it does
OMNIROUTE_API_KEY	bearer token fallback for --bearer (checked first unless --bearer is passed)
TOKTOP_BEARER	bearer token fallback for --bearer (checked after OMNIROUTE_API_KEY)
TOKTOP_SSH_PASSWORD	ssh password for headless runs; otherwise an interactive prompt
TOKTOP_COLUMNS / TOKTOP_LINES	fixed frame size for --once output (screenshots, capture); must be > 40 / > 20, and a set-but-invalid value aborts with exit code 2
GITHUB_TOKEN	optional; authenticates toktop update's GitHub API calls past the anonymous rate limit
GAUNTLET_HOME	directory holding agents.json (default ~/.gauntlet)
XDG_DATA_HOME	with --opencode-db: directory under which opencode/opencode.db is read (default ~/.local/share)
NO_COLOR	strips terminal styling when set to any value (honored by the renderer)

An explicit --bearer, even empty, wins over its env fallbacks; otherwise the environment is used. Prefer an env var over --bearer for tokens: command-line arguments are visible in process listings to every user on the host, and passing --bearer prints a reminder of that. The token travels only to endpoints named with --add; engines found by port scanning receive no credentials, so a hostile listener on a probed port cannot collect your gateway key. --add URLs must be http:// or https:// with a host and must not embed userinfo (user:pass@); an endpoint that needs the key is attached as toktop --add http://127.0.0.1:20128 with the token in the environment. --ingest must be host:port (empty would bind every interface on an ephemeral port and is rejected). Unknown TOKTOP_* variables are reported at startup, so a typo fails loudly instead of doing nothing. Out-of-range flag values (--interval 0, negative --probe, --frames < 1 with --once, a malformed --add or --ingest) abort with exit code 2 instead of being silently adjusted; so do out-of-range TOKTOP_COLUMNS / TOKTOP_LINES when --once renders.

Build & test

make help                          # every task, one line each
make build                         # host binary, version-stamped
make demo                          # build, then run the simulated fleet
make test                          # all tests, -race -shuffle=on
make pr                            # every PR merge gate except the OS matrix
make test-pkg PKG=./internal/ui    # one package while iterating
make test-pkg PKG=./internal/core RUN=TestSanitizeTextPreservesUTF8

Cross-compiles (no cgo anywhere); make test-dist is the same flags the release uses (-trimpath -buildvcs=false -mod=readonly -buildmode=pie):

make test-dist VERSION=x.y.z    # every release platform into dist/

See CONTRIBUTING.md for prerequisites, the edit-test loop, and what CI runs, and docs/THREAT_MODEL.md for the attack surface, what toktop trusts, and the mitigations already in place.

Releases: push a tag v* and GitHub Actions attaches binaries for linux/amd64, linux/arm64, darwin/amd64, darwin/arm64, windows/amd64 and windows/arm64, plus a CycloneDX SBOM of every dependency (make sbom). Versions are 0.x: the CLI, the ingest /v1/events body, and the agentusage Go API may change without a major bump. Consumer-facing notes live in CHANGELOG.md. CI runs govulncheck on every push; Dependabot keeps go modules and workflow actions current.