CrabTalk
[ Open source · Rust ]

Just the daemon.

CrabTalk is an open-source agent daemon written in Rust. Five megabytes, zero children, nothing hidden — just the plumbing every agent app rebuilds, written once and shared.

/* One line, then talk to it … */

$ curl -sSL https://crabtalk.ai/install | sh
Read the docs GitHub
Cross-section of the CrabTalk daemon: one long-lived process with clients attaching from outside and the model across a clean boundary.
Architecture. One long-lived process in the centre; clients — CLI, bots, scheduled jobs — attach from outside; the model lives across a clean boundary in crabllm.
$BINARY_SIZE
5 MB
$CHILD_PROCS
zero
$LANGUAGE
Rust
[ Attach, and speak ]

One command.

One command launches the daemon and drops you into a session. No server to provision, no config to author, no dashboard to sign into. The binary is five megabytes; it holds nothing back.

[ Five pillars ]

What the daemon guarantees

Crashes don’t propagate. Topics route themselves. Memory stays precise. Coordination scales. Each one a pillar — drawn here as a live plate.

Fig. 01 · Pillar 01 / 05

Crash isolation

Source · spec · daemon

In most AI systems, one tool crashes and the whole chat crashes with it.

CrabTalk runs every tool, every command, every extension in its own space. One breaks; the rest keep going.

Fig. 01. Broken parts, stable core. A central body holds; satellites move on their own orbits, breaking apart and re-forming without disturbing the centre.

span 8 · landscape · 1 of 5
Fig. 02 · Pillar 02 / 05

No sessions

Source · spec · conversations

Fig. 02. Topic lanes. Messages travel along their own subject channels — work in the work lane, home in the home lane — without per-window juggling.

Other AI makes you open a new chat per subject.

CrabTalk sorts by topic itself — work stays in the work channel, home switches itself, no window-juggling.

span 4 · portrait · 2 of 5
Fig. 03 · Pillar 03 / 05

Precise memory

Source · spec · memory

Fig. 03. Indexed retrieval. A bound archive built on the same internals engineers trust for their databases — fast lookup that doesn’t slow as it grows.

AI memory usually forgets fast, or slows to a crawl as it grows.

We ported the internals of SQLite to the recall layer — accurate, fast to query, compact on disk.

span 4 · portrait · 3 of 5
Fig. 04 · Pillar 04 / 05

Orchestration at scale

Source · spec · dispatch

Fig. 04. At scale. Hundreds of agents in coordinated motion — a conductor facing a full orchestra, each section playing its own line on cue.

Other systems may cap around 10 concurrent agents.

CrabTalk is designed to coordinate 10,000. A conductor facing a full orchestra, not a small ensemble.

span 8 · compact · 4 of 5
Fig. 05 · Pillar 05 / 05

Guest agents

Source · spec · multi-agent

Fig. 05. One conversation, two voices. A guest agent speaks directly into the shared history — tagged, not delegated — then the primary resumes on the next turn.

Sometimes a second agent should speak into an ongoing conversation — a reviewer, a translator, an expert.

A guest reads the shared history, adds a tagged message, and steps back. The primary picks up where it left off.

span 12 · banner · 5 of 5

Just the daemon.

5 MB · zero children

Read the spec github.com/crabtalk