Memory consolidation skill for Claude Code. Scans session transcripts for corrections, decisions, preferences, and patterns, then merges findings into persistent memory files. Auto-triggers via native Stop hook every 24hrs. Inspired by how sleep consolidates human memory.
Automatically fix skill problems (safe preview first). Also use with --prune to find and remove broken symlinks, empty directories, and orphaned skills. Trigger with '/janitor-fix'.
Show whether each skill is earning its context-window cost — combined tokens-used view sorted by waste. Use when the user asks 'are my skills worth it', 'what's my context budget', 'which skills are dead weight', or wants to audit skill value, token cost, or usage. Trigger with '/janitor-value'.
Memory-agnostic background consolidation — reviews sessions, scores candidates, promotes durable insights to the active memory backend (built-in MEMORY.md or Holographic). Three-phase (Light/Deep/Condensation/REM). Run via cron or manually.
Audits all memory systems in a Perseus environment, identifies stale or superseded stores (Mem0, Mempalace, old vector DBs, garbled federation artifacts), and deletes them. Keeps Perseus Vault (the live system) as the single source of truth. Run periodically or when upgrading Perseus versions.
Natural-language editor for .claude/quality/schedule-config.json. Users talk about maintenance scheduling in everyday language ("run audit twice a week", "turn off deep daydream", "switch dispatcher to 08:30") — this skill parses intent, validates against the schema, shows a concrete diff, asks for confirmation, and…
Memory health CLI that diagnoses bloated MEMORY.md files, splits oversized sections into domain files, archives stale daily logs, and runs continuous health monitoring. Use when MEMORY.md exceeds 10KB, context windows overflow from memory bloat, or agent sessions slow down from loading large memory files. Pure bash…
Review, diagnose, and refine existing agent skills as living systems. Use when Codex needs to explain why an existing skill misfires, assess its architecture or evolution, apply an explicitly requested minimal improvement, recheck prior findings, or audit drift across source, published, and installed copies.
Normalize one existing release-note entry for consistent headings and links. Use when the user asks to clean a release note without changing its facts.
A periodic cleanup framework for an AI memory store: a collection of saved facts used to retain knowledge between sessions. It sorts, combines, reviews, updates, and reorganizes facts in a chosen area.
Regelmäßiger Pflegedurchlauf für ein bereits veröffentlichtes GitHub-Repository (Stufe 1, günstig und oft wiederholbar): zuerst alle Distributionsflächen des Projekts ermitteln (npm, PyPI, Registries, Marketplaces, Stores, Website) und Änderungen später dorthin spiegeln, dann Topics setzen, Privacy-Gate, Dokumente auf…
Einmaliger Tasksolver+Writer+Maintainer-Durchlauf fuer den Projektordner, in dem die Session gerade sowieso gearbeitet hat (inkl. Unterordner) — kein systemweiter Sweep. Loest offene triviale/autonome Register-Punkte, aktualisiert Dokumentation auf den gemessenen Ist-Stand (im Projekt UND in der Root der Pipeline, in…
Use when assessing tire condition — measuring tread depth, interpreting wear patterns, calculating remaining safe life, and determining when tires need replacement based on safety and legal standards.
Diagnose a failing service or system issue on a Debian/Ubuntu Linux machine with a hypothesis-driven loop that gathers evidence from journalctl and systemctl, finds patterns, tests a hypothesis, fixes, and verifies. Use when something is broken and you need root cause, not a guess.
Routine maintenance pass on a Debian/Ubuntu Linux machine that updates and upgrades packages, autoremoves, cleans cache, scans logs, checks disk and services, and verifies backups. Asks before destructive steps. Use for regular upkeep.
Analyze and optimize performance on a Debian/Ubuntu Linux system by assessing CPU, memory, disk, network, and processes, finding the bottleneck, then applying targeted tuning with before-and-after measurement. Use when a machine is slow and you need to find why.
Self-maintenance discipline for any repo — start-of-day bootstrap, end-of-day wrap, and metrics. Grounds the session in GitHub truth (open PRs, issues, CI, DORA) BEFORE starting new work, finishes in-flight work first, keeps issues/PRs/worktrees clean, records DORA, and stops maintenance from starving behind features.…
Produce "the weekly-seedling" — this repo's weekly newsletter on the state of the art in its own domain. GitHub-first research (repos/papers/techniques relevant to THIS repo) run through a default-cut editorial gate, rendered as a self-contained HTML issue and filed to the board. Each item ends in an…