hatchet

Inspect Hatchet workflows and failed runs on the Railway tsmesh deployment via CLI.

Updated May 11, 2026
One-click install
npx skills add https://github.com/mmnavarr/harness --skill hatchet-mmnavarr
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: hatchet
Source: https://github.com/mmnavarr/harness/tree/main/skills/hatchet
Command: npx skills add https://github.com/mmnavarr/harness --skill hatchet-mmnavarr

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? Connecting the Hatchet CLI to the Alliance Railway deployment is error-prone: tokens embed stale internal URLs, there is no login command, and local workers confuse the API route with the gRPC endpoint. This Skill provides the exact profile setup, token handling, and read-only inspection workflow for the railway-monorepo instance. ## Core Features & Use Cases - Safe CLI profile setup: Create the railway-monorepo profile using fnox/mise token injection with explicit API and TLS overrides, without ever printing the token. - Read-only run inspection: List workflows, list failed runs, inspect individual runs, and launch the Hatchet TUI against http://railway-monorepo:11744. - Troubleshooting guidance: Diagnose stale token-derived URLs, tsmesh/DNS failures, and the distinction between the UI/API route and the hardcoded gRPC endpoint hatchet.railway.internal:7077 used by the BullQueue worker. - Use Case: You need to check why jobs failed last week on the Alliance Railway Hatchet instance. The Skill walks you through verifying the route health, injecting the token via fnox, and running hatchet runs list with the correct overrides. ## Quick Start Ask the assistant to list failed Hatchet runs from the last 7 days on the railway-monorepo profile using the fnox-injected token and the tsmesh API overrides.

Frequently Asked Questions about hatchet

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I connect the Hatchet CLI to a Railway deployment?▼

Create a profile with hatchet profile add using a token injected via fnox and mise, passing HATCHET_CLIENT_SERVER_URL=http://railway-monorepo:11744 and HATCHET_CLIENT_TLS_STRATEGY=none. There is no hatchet login command; profile add validates the token before saving.

How do I list failed Hatchet runs from the command line?▼

Run hatchet runs list with the railway-monorepo profile, --status FAILED, --since 7d, and --output json, prefixed with the HATCHET_CLIENT_SERVER_URL and HATCHET_CLIENT_TLS_STRATEGY=none overrides so the CLI uses the tsmesh API route.

Why does hatchet profile add connect to the wrong URL?▼

Hatchet tokens embed claims like server_url and grpc_broadcast_address, which may be stale or internal. Override them with HATCHET_CLIENT_SERVER_URL=http://railway-monorepo:11744 during profile creation and every command instead of decoding or printing the token.

Can a local Hatchet worker use the railway-monorepo:11744 endpoint?▼

No. Port 11744 is the UI/API route; workers need the engine gRPC endpoint, hardcoded in this repo as hatchet.railway.internal:7077. Local workers require a tunnel preserving that host and port, or must run inside Railway.

Why does hatchet runs list work but local worker startup fail?▼

CLI inspection uses the REST API route while workers connect over gRPC to hatchet.railway.internal:7077, which is only reachable inside Railway's private network unless tunneled. Also confirm HATCHET_CLIENT_TOKEN is injected via fnox before starting BullQueue.