redskilled

Operates the redskilled host daemon and writes debug dossiers for issue Workers.

20|2|Updated May 16, 2026
One-click install
npx skills add https://github.com/reddb-io/red-skills --skill redskilled-reddb-io
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: redskilled
Source: https://github.com/reddb-io/red-skills/tree/main/plugins/dev/skills/engineering/redskilled
Command: npx skills add https://github.com/reddb-io/red-skills --skill redskilled-reddb-io

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires @reddb-io/red-skills.

What problem does it solve? Operators running the per-machine redskilled execution daemon need a single guided workflow to inspect status, provision the host, configure machine policy, restart safely, and diagnose why a Worker on a given issue stalled or never started. ## Core Features & Use Cases - Daemon lifecycle operation: Read provisioning audits and host state, provision the daemon home and OS service, configure worker/memory/validation ceilings and GitHub credential profiles in ~/.red/config.yaml, and restart through the idempotent provisioning command. - Self-upgrading daemon guidance: Distinguishes a daemon that is merely behind the registry (wait one replace interval) from a wedged daemon (stop and re-provision), preventing harmful manual restarts. - Issue-keyed Worker debugging: Resolves an issue number to every Worker that touched it via MCP tools and log lanes, then writes a self-serve debug dossier to .red/tmp/diagnostics/ with fiche, event sequence, log excerpts, diagnosis, and next moves. - Use Case: An operator runs /redskilled 3351 to learn why the Worker on issue 3351 never started, and receives a dossier showing the daemon's refused-over-worker-ceiling admission verdict plus the exact recovery commands to run. ## Quick Start Ask the assistant to operate the redskilled daemon with no argument, or pass an issue number like 3351 to debug that issue's Worker and produce a diagnostic dossier.

Frequently Asked Questions about redskilled

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

FAQPage Schema
How do I debug a redskilled Worker by issue number?▼

Pass the issue number as the skill argument, with or without a leading hash. The skill resolves the issue to Workers via the rs_dev MCP status and claim_status tools, falls back to scanning .red/tmp/workers/ lanes, and writes a debug dossier to .red/tmp/diagnostics/.

How do I provision the redskilled daemon on a fresh machine?▼

Run the red-skills-redskilled provision command with --workspace host via npx. It creates ~/.red/redskilled/ and the initial config template, installs the always-on OS service, starts the daemon, and prints the audit.

Should I stop a redskilled daemon that is behind the latest published version?▼

No. The daemon upgrades itself by re-reading the registry every fifteen minutes, so a trailing daemon_version is normal. Use stop only for a genuinely wedged daemon whose socket no longer answers.

Where does redskilled store Worker and daemon logs?▼

Worker logs live in .red/tmp/workers/<worker-id>/worker.log.toonl and the host-scoped daemon log is ~/.red/redskilled/redskilled.log.toonl. All are TOONL format, readable with the tq tool using the nearest segment header.

Can a project configure GitHub credentials for redskilled Workers?▼

Projects may only name a public credential_profile in tracked config. Tokens, PEM files, App ids, and installation ids stay in the host ~/.red/config.yaml under plugins.dev.redskilled.github_profiles, and are never exposed to clients or Workers.

What are the limitations of the redskilled debug entry?▼

The debug entry is strictly read-only: it recommends recovery verbs like hitl_resolve or claim_release in the dossier but never executes them. There is also no issue-keyed resolver tool, so Worker resolution is a client-side filter plus lane scan.