caveman
claudeTalk like caveman. Cut filler. Keep technical details explicit.
22 skills · 1 command · 3 agents · hooks
Bring your organization's context lake into your coding agent — fused org search across code, decisions, tickets, PRs, docs, and people; graph-powered code navigation (callers, blast radius, tests-for); standing org rules and durable memories, served over the bonez MCP.
Plugin bonez from marketplace bonez in bonez-io/ai-plugin. Commands read from the repo's manifest — 3 days ago.
This plugin runs code on your machine — its hooks fire automatically on events like tool use. The exact commands are listed below; read them before installing.
citing-bonez-sourcesCite and dereference Bonez results correctly. Use whenever you present or act on a finding that came from a Bonez tool, when a result carries a ~hex handle, when the user asks where a claim came from, or when you need the full record behind a search result.
creating-a-pluginCreate, extend, test, bundle and hand over a Bonez plugin (a "package" that adds tools to Bonez agents). Use when the user wants to write a Bonez plugin, package, extension or new agent tool, add a tool to one, build or hash one, or asks how a plugin gets onto a Bonez box.
creating-an-agentCreate or change a Bonez agent from this session, one a person starts, one that runs on a schedule, or one that wakes on an event (a Linear issue, a GitHub pull request). Use when the user asks to create, make, set up, schedule, edit, pause or retire a Bonez agent ("run this every night", "make an a
finding-prior-artSearch what the org already knows before you build or decide. Use BEFORE you write non-trivial code, design an approach or choose a library, when the user asks "have we done this before", "why is X like this" or "has anyone hit this error", when a decision smells like it was already made once, or be
impact-analysisMeasure the blast radius before a change lands. Use BEFORE you change, rename or delete a function, type, file or public contract that other code may use, when asked "what breaks if", "who calls this", "is it safe to change or delete this" or "what tests cover this", before a refactor of a shared sy
querying-the-graphTraverse the org graph with BGQ via graph_query. Use when a question is about relationships or structure (what calls, imports, tests or depends on X, what is connected to a ticket or PR, which code touches a table), when you would otherwise grep across repos to trace callers or dependencies, or when
rememberingSave durable facts into the org's memory with graph_write. Use when you learn something a teammate would want next time (a convention, a decision and its why, a non-obvious gotcha, a correction the user made), when the user says "remember", "from now on", "always" or "never", or when deciding whethe
reviewing-with-org-rulesReview code against the org's standing rules. Use when asked to review a PR, diff or change in a Bonez-connected org, when asked whether code follows the org's conventions, before you tell the user a non-trivial change is ready, and before approving, merging or signing off on anything non-trivial. P
session-contextOrient in the org's Bonez graph. Use at the start of a task in an unfamiliar repo, when resuming work, or in a new session in a Bonez-connected org, when you need the org's standing rules or a map of what the graph holds, when unsure which conventions apply, or before a call an org rule may already
suggesting-agentsPropose a Bonez agent, once, when the work will repeat. Use when the user fixes a bug that had come back before or says "again", "keeps happening", "regression", "this broke last month" or "make sure this never breaks again" (propose a nightly agent that re-checks the fix), when they do something pe
using-the-tool-lakeReach the org's connected vendor systems (GitHub, GitLab, Jira, Linear, Monday, Sentry, Slack, Discord, databases) with tool_search, vendor_operation and vendor_operation_status, as the signed-in user. Use when a task needs a ticket, PR or merge request, CI run, error, thread or table from one of th
who-owns-whatFind people (owners, experts, reviewers) through the org graph. Use when asked who owns, maintains, wrote, changed, decided or knows something ("who owns billing", "who last changed this", "who should review this", "who do I ask"), when you must pick a reviewer or assignee, or when routing work to a
/bonez:agentsList your Bonez agents and their latest runs, or show one agent's recent runs
/bonez:connectConnect this plugin to your own Bonez server and sign in — OAuth or an API key
/bonez:contextOrient in your Bonez graph — what it holds, the standing rules in effect, and (optionally) what it knows about a repo or topic
/bonez:new-pluginScaffold, test, bundle and hash a new Bonez plugin (a package that adds tools to Bonez agents), then publish it to your Bonez server (or print the manual handoff if that fails)
/bonez:searchSearch your Bonez graph — code, docs, tickets, PRs, conversations, people — in one query
bonezconnects to: ${user_config.bonez_url}/mcp
bash "${CLAUDE_PLUGIN_ROOT}/hooks/gate-write.sh"
node "${CLAUDE_PLUGIN_ROOT}/bin/bonez-session-sync.mjs" hook claude session-end
node "${CLAUDE_PLUGIN_ROOT}/bin/bonez-session-sync.mjs" hook claude session-start
node "${CLAUDE_PLUGIN_ROOT}/bin/bonez-recall-hook.mjs" claude
node "${CLAUDE_PLUGIN_ROOT}/bin/bonez-recall-hook.mjs" claude
node "${CLAUDE_PLUGIN_ROOT}/bin/bonez-recall-hook.mjs" claude
node "${CLAUDE_PLUGIN_ROOT}/bin/bonez-recall-hook.mjs" claude
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
AI-powered codebase understanding — analyze, visualize, and explain any project
9 skills · 10 agents · hooks
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
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.