caveman
claudeTalk like caveman. Cut filler. Keep technical details explicit.
22 skills · 1 command · 3 agents · hooks
Brain context injection, session capture, telemetry, slash commands, skills, and the agenticos MCP server for an AgenticOS vault.
Plugin agenticos from marketplace agenticos-workbench in zzoretich/UniDeX-Agent-Harness. Commands read from the repo's manifest — today.
This plugin runs code on your machine — its hooks fire automatically on events like tool use and it starts local MCP servers. The exact commands are listed below; read them before installing.
browserUse the UniDeX browser, the browser inside the UniDeX app that the user watches beside this session, to open a web page or a site, search the web, read documentation, check localhost, or preview what you built. Use when the user says "the browser", "open this page", "look at this site", "check local
costSession and month-to-date cost from transcripts — cost completed sessions, read the anchored budget, explain where the money went. Only active when the cost module is enabled (cost.enabled in agenticos.json after `aos cost enable`); otherwise explain how to enable it. Use when the user asks "how muc
cross-reviewHarden a plan with an independent review by the other provider (Claude Code and Codex review each other), then optionally build it and have the provider that did not build inspect the final diff in a fresh session. Adapted from claudex-loop. Use when the user says "cross-review this plan", "cross-re
feedback-reviewBatch-review the draft feedback rules captured automatically at session end by the correction detector and by wrap_session corrections — approve, reject, or flag conflicts in one pass. Use when the user says "review feedback drafts", "approve feedback drafts", "feedback drafts", "pending feedback ru
graphStructure questions about the vault through its knowledge graph — how two notes or projects connect, what sits around a note, idea or project, which notes are hubs, which clusters exist. Use when the user asks "how is X related to Y", "what links to X", "what's around this project", "map my notes on
handoffRecommend who should handle a task and run one scoped handoff when asked. Stay with the current agent, get a second opinion from the other provider, investigate a blocker, or delegate a bounded piece to a suitable model on Claude Code or Codex. Adapted from claudex-route. Use when the user asks "who
persona-flag-closerBatch-review your Chief of Staff agent's pending proposals, STATE.md flags, and new silent duty failures — deterministic re-verification via per-proposal recheck recipes, one HTML digest, one decision pass for ideas and flags, with exact instruction changes sent to the Proposals UI, and one commit p
persona-sitrepMorning work-state sitrep from your Chief of Staff agent — what moved across repos and planning phases, what's stalled, what's pending approval, with exactly ONE recommended action. Use when the user says "sitrep", "/agenticos:persona-sitrep", "what's next", "what should I work on", "work status", o
recallRanked, recency-weighted recall over the whole vault — memories, patterns, daily notes, and the persona journal. Use when the user asks "when did I last touch X", "what did we decide about Y", "have we seen this before", "search my memory", "search the vault", "/recall", or at the start of any non-t
wrapIn-session memory extraction at the end of a Claude Code session — review the conversation, then call the wrap_session tool once so memories, MEMORY.md lines, SESSION.md sections and feedback drafts are written through the same writer the background auto-wrap uses. Use during /wrap, when SESSION.md
/agenticos:agentsEvery Claude Code and Codex agent in one list, and share each host's own agents with the other — list, sync now, stop or resume sharing one agent, reset an edited copy
/agenticos:aosUniDeX maintenance — doctor, status, provider, persona, config (runs the aos launcher)
/agenticos:ask-brainAsk the vault a question — a script assembles the relevant memories, you answer from them (or --local lets the local provider answer)
/agenticos:brainShow current brain state — BRAIN.md, SESSION.md, and the most recent daily notes
/agenticos:compressCompress a large file or piped text into a distillate before reading it in full
/agenticos:consolidate-memoryFind duplicate or overlapping memories and patterns and propose merges — a review draft only, sources are never edited
/agenticos:costCost completed sessions from their transcripts and patch runs.jsonl (only when the cost module is enabled)
/agenticos:feedbackCapture user feedback as a permanent rule in brain/memory/feedback/ and index it in MEMORY.md
/agenticos:notificationsRead and manage the vault's notifications (the Workbench Notifications tab) — show unread items, mark read, archive; and how any agent posts one with `aos notify post`
/agenticos:patternCreate or extend a decision-heuristic note in brain/patterns/
/agenticos:projectCapture project context as a memory in brain/memory/projects/ (active projects are compiled into BRAIN.md)
/agenticos:proposeFile a proposal in persona/proposals (the Workbench Proposals tab) in the standard format, with its HTML page and a link to it
/agenticos:reflect-weekWeekly reflection from the last 7 daily notes and git history, written to brain/reflections/
/agenticos:rememberAppend a note to SESSION.md "Things to Remember"; tag #promote to make it permanent at /wrap
/agenticos:routinesManage the vault's recurring actions (brain/routines/*.md) — list, sync the OS schedules, run one now, enable or disable, next fire times; see the routines each session host owns (Codex Automations, Claude Code cloud routines)
/agenticos:scanRefresh the dashboard caches — full vault scan, BRAIN.md compile, recall index
/agenticos:skillsEvery Claude Code and Codex skill in one list, and share each host's own skills with the other — list, sync now, stop or resume sharing one skill, reset an edited copy
/agenticos:standupDid / Doing / Blockers standup from recent daily notes, git and telemetry, appended to today's daily note
/agenticos:teamShow and run the vault's agent teams (the Workbench Agent Teams tab) - boards, pending gates and live runs; record a gate the user decides; and how a lead dispatches seats and every member posts with `aos team`
/agenticos:todoAdd a todo to TODO.md in Obsidian Tasks syntax (due date, priority, #tags); the Workbench To-Do tab shows it
/agenticos:wrapSession-end protocol — promote #promote items, extract this session's memories in-session (wrap_session), summarize into today's daily note, reset SESSION.md
agenticosruns: sh ${CLAUDE_PLUGIN_ROOT}/bin/aos mcp-server
sh "${CLAUDE_PLUGIN_ROOT}/bin/aos" telemetry-hook
sh "${CLAUDE_PLUGIN_ROOT}/bin/aos" update-notice
sh "${CLAUDE_PLUGIN_ROOT}/bin/aos" persona-watchdog
sh "${CLAUDE_PLUGIN_ROOT}/bin/aos" reconcile-sessions
sh "${CLAUDE_PLUGIN_ROOT}/bin/aos" inject-context
sh "${CLAUDE_PLUGIN_ROOT}/bin/aos" instruction-guard
sh "${CLAUDE_PLUGIN_ROOT}/bin/aos" telemetry-hook
sh "${CLAUDE_PLUGIN_ROOT}/bin/aos" update-session
sh "${CLAUDE_PLUGIN_ROOT}/bin/aos" heartbeat-writer
sh "${CLAUDE_PLUGIN_ROOT}/bin/aos" reconcile-sessions
sh "${CLAUDE_PLUGIN_ROOT}/bin/aos" telemetry-hook
sh "${CLAUDE_PLUGIN_ROOT}/bin/aos" auto-cost
sh "${CLAUDE_PLUGIN_ROOT}/bin/aos" heartbeat-writer
sh "${CLAUDE_PLUGIN_ROOT}/bin/aos" auto-wrap
sh "${CLAUDE_PLUGIN_ROOT}/bin/aos" scan-vault --quiet
sh "${CLAUDE_PLUGIN_ROOT}/bin/aos" skills-sync
Talk like caveman. Cut filler. Keep technical details explicit.
22 skills · 1 command · 3 agents · hooks
Persistent memory for Claude Code: the next session starts where the last one left off.
29 skills · 1 MCP server · hooks
tasteskill 2.5.1: professional company websites at the figma.com and x.ai level, with one section per screen, real photographs, real product UI, effects kept for tech and gaming pages, and a real-browser checker and finish gate.
1 skill
Open-source AI job search that runs locally in your AI coding CLI: scan job portals, evaluate listings into a structured A-H report with a 1-5 score, tailor your CV, track applications
5 skills · hooks
Cross-session memory and token savings for coding agents.
6 skills · 1 agent · 1 MCP server · hooks
Package any website or local web build into a lightweight desktop app with Pake CLI.
1 skill
This page updates itself from your repo. Add the badge to your README, or get it Featured to pin it above the organic results with a followed link to your site.