legible.sh

Infrastructure legible enough for an agent to use on the first try.

Ten zero-dependency primitives, each doing for a different verb what ntfy.sh did for notifications: the URL is the API, no signup, the unguessable name is the capability, and the whole tool fits in a prompt. Built for people whose agents are smart but infrastructurally naked.

# block a deploy on a human tap:
curl -d "Deploy v2 to prod?" "https://gate.legible.sh/ops-x7k2q?wait=300" && ./deploy.sh

# give every agent loop an emergency brake:
curl -fs https://bigred.legible.sh/ok/fleet-7 || exit 1

# count something; watch it as a live chart:
curl -d +1 https://tally.legible.sh/myrun/tasks-done

🚪 gate askntfy tells you things. gate asks you things.gate.legible.sh github

🔴 bigred stopThe big red button for your agent fleet.bigred.legible.sh github

🛤 trail logThe flight recorder for agent runs.trail.legible.sh github

🖤 slate shareThe blackboard from the multi-agent papers, as a URL.slate.legible.sh github

📮 relay queueThe work queue your agents can provision themselves.relay.legible.sh github

🔒 mutex lockflock(1) for agents that live on different machines.mutex.legible.sh github

🗳 quorum decideCoordination for agents that do not share a parent process.quorum.legible.sh github

⛽ meter capThe kill-brake for agent spend. 429-as-a-service.meter.legible.sh github

📦 stash moventfy moves signals. stash moves bytes.stash.legible.sh github

📈 tally countStatHat reborn as ntfy.tally.legible.sh github