ShellYard helps teams capture the terminal output, diagnostics, API results, and context behind a fix — then turn it into documentation that doesn't go stale.
The problem
The runbook says what to do when the API times out, but it was written for the system as it was 18 months ago. The 3 a.m. on-call engineer reads it, realizes half the commands no longer work, improvises a fix into a Slack thread, and the next on-call hits the same problem from the same starting point. Operational memory leaks out faster than you can write it down.
ShellYard remembers what your terminal just showed. Highlight the relevant lines, attach them to a runbook draft, and you have a documented troubleshooting step grounded in real output — not a paraphrased recollection two days later. The same is true for HTTP responses, database query results, packet captures, and config diffs.
Attach the terminal session, the relevant tool output, and a one-line description of the problem. Magellan sees everything ShellYard captured during the session — the connection, the credentials path, the diagnostics you ran, the response payloads — and pulls them in as grounding for whatever you ask next.
Magellan drafts a runbook with the steps you took, the commands you ran, the output that mattered, and a suggested rollback. You read it, edit it, and save it. The work of writing the runbook drops from 30 minutes of dread to 5 minutes of editing. BYO Anthropic / OpenAI / Gemini / Ollama key; operators always run the actual command.
Runbooks, notes, IPAM records, and incident reports live in the same Space as the connections, credentials, and tools they reference. Open the client Space and the runbooks are there — not in a separate documentation portal three browser tabs away. Version history on every save (90 days on Team, unlimited on Enterprise).
Incident report scaffolds capture what happened, what was checked, what changed, and how it was resolved. Timeline, evidence sections (terminal output, HTTP response, DB query, log snippets, config diff), root cause, resolution, follow-ups, customer-facing summary. Not a replacement for PagerDuty or Jira — it's the documentation half of post-incident work, where it actually belongs.
When the only person who knows how to recover the cluster is on PTO, your team has a problem. Runbooks built from real troubleshooting sessions — not theoretical write-ups — close that gap. Team and Enterprise plans add Shared Spaces, version history, and audit-log export so the institutional memory is something the whole team can see and trust.
Pro adds Shared Spaces with one teammate. Team unlocks shared Documents and IPAM, audit export, and IT Glue / Hudu importers. Enterprise adds unlimited version history per Space.