القائمة
Run note quality maintenance — invoke /inspect to diagnose vault metadata, then /repair to fix queued issues. Use when the user wants weekly or periodic Rill vault health maintenance.
التثبيت باستخدام Codex أو Claude انسخ هذا Prompt والصقه في Codex أو Claude أو مساعد آخر ليراجع صفحة Skill ويثبّتها لك.
استنادا إلى تصنيف SOC المهني
Generate today's Daily Note (or one for a specified date) at reports/daily/YYYY-MM-DD.md by aggregating the day's situation across journals, tasks, workspaces, knowledge, activity-log, and any plugin briefing hooks. Fully automated, no interaction. Use when the user asks for a daily briefing, situational analysis, or "what's going on today/yesterday".
Ingest a single Twitter/X tweet URL into inbox/tweets/, fetch its content via fetch-tweet.sh, and generate an organized version with frontmatter and Markdown body in _organized/. Combines `rill clip` with the tweet branch of /distill Phase 2 in one shot. Use when the user supplies a single tweet URL to capture.
Quantitatively evaluate the search quality and efficiency of the Rill knowledge base via natural exploration vs Deep Path ground truth (ADR-050). Run monthly. Computes precision, recall, metadata_contribution, token cost. Use when the user wants to verify whether note structure actually works in practice.
Start (or resume) a Rill workspace from a theme, journal file, or workspace id. Automatically collects related vault context and creates a workspace/{id}/ directory for an interactive Deep Think session. Use when the user wants to begin or continue a workspace-scoped exploration.
Diagnose Rill PKM taxonomy health, metadata accuracy, and file integrity. Run weekly. Detects issues and appends mismatched files to knowledge/.refresh-queue for repair by /repair.
Run the daily morning routine — invoke /briefing (Daily Note) followed by /newsletter (research report) inline within the current session. Use when the user starts the day or explicitly asks for the morning routine.
| name | maintain |
| description | Run note quality maintenance — invoke /inspect to diagnose vault metadata, then /repair to fix queued issues. Use when the user wants weekly or periodic Rill vault health maintenance. |
Conduct all conversation with the user — and write all generated output — in the language defined by .claude/rules/personal-language.md (or the user's input language if that file is absent). Follow the language rules in full — exceptions and translation quality are defined in the Language Rules of .claude/rules/rill-core.md and the vault's personal-*.md overrides, never restated per skill. The English instructions below are for skill clarity, not for output style.
Tool references in this skill (
sub-agent, "skill invocation") describe intent, not Claude-specific tool calls. Each harness should map them to its native equivalent — Claude Code uses its built-in Skill / Agent tools as named; Codex CLI uses its own skill invocation mechanism (mention / implicit //skills) and sub-agent mechanism as appropriate.
Runs note quality inspection and repair sequentially. Diagnoses issues with /inspect, then fixes them with /repair if any are found.
$ARGUMENTS — None
Invoke the inspect skill (via the harness's skill mechanism).
/inspect diagnoses taxonomy health + metadata accuracy + consistency, and adds problematic files to knowledge/.refresh-queue.
Read knowledge/.refresh-queue and check the count.
Invoke the repair skill (via the harness's skill mechanism).
/repair reads files from .refresh-queue and batch-normalizes tags/mentions/type.
/inspect and /repair use sub-agents internally, but are not as heavy as /distill. The combined context consumption of the 2 skills is manageable (~50-80K tokens). To avoid the process startup overhead of spawning a new harness session, they are run sequentially inline within the current session.
