From lazycortex-core
Show lazycortex-core purpose and a one-line summary of each skill it ships
How this command is triggered — by the user, by Claude, or both
Slash command
/lazycortex-core:lazy-core.helpThe summary Claude sees in its command listing — used to decide when to auto-load this command
Output the block below verbatim to the user. Do not summarize, rephrase, or add commentary. Do not invoke any tools. Do not log this run. --- **lazycortex-core** — opinionated hygiene layer for Claude Code configs. Audits what's loading into context, slims oversized rules, flags secrets/PII/paths before public commits, and batches MCP tool permissions. **Skills** (invoke as `/<name>` or via Skill tool): - `lazy-core.agent-models` — interactively assign model tiers (haiku/sonnet/opus/inherit) to every dispatchable subagent missing from `lazy.settings.json`; routes each entry to its struc...
Output the block below verbatim to the user. Do not summarize, rephrase, or add commentary. Do not invoke any tools. Do not log this run.
lazycortex-core — opinionated hygiene layer for Claude Code configs. Audits what's loading into context, slims oversized rules, flags secrets/PII/paths before public commits, and batches MCP tool permissions.
Skills (invoke as /<name> or via Skill tool):
lazy-core.agent-models — interactively assign model tiers (haiku/sonnet/opus/inherit) to every dispatchable subagent missing from lazy.settings.json; routes each entry to its structurally-correct scope.lazy-core.audit — read-only report of what gets loaded into context at startup, by category and size.lazy-core.doctor — health check across rules, agents, skills, commands, settings, memory, hooks, CLAUDE.md. Delegates to sibling audits.lazy-core.install — bootstrap the plugin into a project (or globally); copies hygiene + security rule templates. Idempotent.lazy-core.optimize — slim oversized rule files, move reference material to agents, fix global-vs-local settings leakage.lazy-core.setup — meta-installer that runs every enabled lazycortex plugin's install skill in dependency order. Idempotent.lazy-expert.cancel-job — cancel an expert job by removing its directory. Confirms via AskUserQuestion for non-done jobs.lazy-expert.collect-job — collect the result of a dispatched expert job; returns {status, response}.lazy-expert.dispatch-job — dispatch a job to a named expert queue; returns {job_id, queue_path}.lazy-expert.list-jobs — list expert queue jobs, optionally filtered by expert name or status.lazy-guard.allow-mcp — add all tools of one or more MCP servers to permissions.allow at the correct scope.lazy-guard.check-public — scan a repo (or subtree) for leaked secrets, PII, internal hostnames, hardcoded local paths. Respects .guard-waivers.json.lazy-repo.mark-public — walk a repo through the check-public audit, create a waivers file, optionally flip GitHub visibility.lazy-log.clean — interactive housekeeping for .logs/claude/; classifies each run-log directory against the live skill/agent/command name set, surfaces orphans and unnamed clusters, and applies merge/delete/leave per user confirmation.lazy-routine.register — register a named routine in lazy-core.runtime; used by plugin install skills.lazy-routine.unregister — remove a named routine from lazy-core.runtime; protects the built-in expert-pump.Commands (invoke as /<name>):
lazy-core.checkup — single entry point: runs every read-only audit/doctor (consumer + author trios), merges findings into one per-plugin table, then prompts for the mutating fix-flow to run.lazy-core.help — this listing.Agents (dispatched via Agent(subagent_type: "lazycortex-core:<name>")):
lazy-log.bullets — drafts user-facing changelog bullet blocks from recent distilled entries.lazy-log.distill — rolls commits in .logs/commits.jsonl into themed changelog entries in .logs/changelog.md.lazy-log.recall — searches change history across .logs/changelog.md, per-run log files, .logs/commits.jsonl, and git log; returns ranked results with git SHAs.lazy-log.summary — synthesises a multi-source narrative summary for a given topic or time range.lazy-log.timeline — produces a chronological view of changes for a date range or topic.Rule:
lazy-log.logging — run-logging contract; every skill, agent, and command must log each run to .logs/claude/<name>/YYYY-MM-DD_HH-MM-SS.md.Documentation:
Offline copy at ~/.claude/plugins/cache/.../claude/lazycortex-core/help/.
See README.md in the plugin for full scenarios and examples.
npx claudepluginhub mebius-san/lazy-cortex --plugin lazycortex-core