メニュー
One-time setup for hermit-scribe — appends the Issue Filing block to CLAUDE.md/CLAUDE.local.md and seeds an auto-mode environment entry naming api.github.com as a trusted, always operator-confirmed destination. Run once per project; re-run to refresh after an upgrade.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
Reflect on recent work and propose improvements if patterns are noticed.
Schedules routines via one persistent Monitor subprocess (zero-token skips); CronCreate fallback where Monitor is unavailable. heartbeat-restart stays a CronCreate re-arm anchor.
Initializes the autonomous agent in the current project. Creates the state directory, templates, OPERATOR.md, and config.json. Appends session discipline to CLAUDE.md. Detects installed hermits. Run once per project, like git init.
Morning house brief — live status, overnight anomalies, energy snapshot, pending proposals, and today's priorities. Runs as a daily routine or on demand.
Evening house brief — end-of-day security check, device status, and energy snapshot. Runs as a daily routine at 22:30 or on demand.
Read site deployment logs, server logs, site application/nginx logs, background process logs, and specific deployment logs from Laravel Forge. Includes a triage mode that bundles recent logs with deployment history for rapid incident analysis. Triggers on "show logs", "read logs", "deployment log", "server log", "what happened to the deployment", "triage failing site".
| name | hatch |
| description | One-time setup for hermit-scribe — appends the Issue Filing block to CLAUDE.md/CLAUDE.local.md and seeds an auto-mode environment entry naming api.github.com as a trusted, always operator-confirmed destination. Run once per project; re-run to refresh after an upgrade. |
hermit-scribe is a maintainer utility with a single skill (/hermit-scribe:hermit-scribe) — this hatch only needs to make its GitHub App posting behavior visible to the operator's CLAUDE.md and to Claude Code's auto-mode classifier. There is no config, no routines, no channels.
Check if .claude-code-hermit/ exists in the current project.
AskUserQuestion) "Core hermit isn't set up yet. Run /claude-code-hermit:hatch now?" with options Yes — run now / No — I'll do it later. If yes, invoke /claude-code-hermit:hatch via the Skill tool and stop. If no, stop.Resolve target file: read .claude-code-hermit/state/hatch-options.json. "target": "local" → CLAUDE.local.md; "target": "committed" or the file absent → CLAUDE.md.
Read target_file (a missing file is marker-absent — the append below will create it). Look for the marker <!-- hermit-scribe: Issue Filing -->.
${CLAUDE_PLUGIN_ROOT}/state-templates/CLAUDE-APPEND.md to target_file (create the file if it doesn't exist).Run bun ${CLAUDE_PLUGIN_ROOT}/scripts/automode-env.ts .claude/settings.local.json — always .claude/settings.local.json, regardless of hatch_target: Claude Code's auto-mode classifier reads autoMode config only from local/user scope, never a committed project .claude/settings.json. This names api.github.com (scoped to the configured HERMIT_GH_REPO, or the default gtapps/claude-code-hermit) as a service the hermit posts to only with the operator's in-session confirmation — context for the classifier, not a standing permission grant (filing still goes through the skill's own preview/confirm gate every time). Additive and idempotent; safe to re-run. No prompt needed.
Print: "hermit-scribe active. Issue filing goes through /hermit-scribe:hermit-scribe, always with an in-session preview before posting."