Skill Claude CodeCodex
Explore, clean, and analyze a dataset to answer a real question, with honest handling of uncertainty and bias. Trigger when the user asks to "analyze this data", "what does this dataset show", "explore this CSV", works with pandas/numpy on tabular data, or wants a data-driven answer to a business/research question.…
Skill Claude CodeCodex
Produce clear, accessible charts and dashboards. Trigger when the user asks to "make a chart", "visualize", "plot", "graph", "build a dashboard", or turn numbers/CSV/query results into a picture. Covers chart-type choice, color and accessibility rules, and shipping the result as one self-contained HTML file the user…
Skill Claude CodeCodex
Systematically root-cause a bug instead of guessing. Trigger when the user asks to "debug this", "why is this failing", "it's broken", "reproduce this bug", "find the root cause", or pastes an error/stack trace. Reproduce, isolate, form one hypothesis at a time, fix at the root, then add a regression test — never…
Skill Claude CodeCodex
Write and improve Dockerfiles that build small, cache-friendly, secure images. Trigger when the user asks to "write a Dockerfile", "containerize this", "shrink my image", "why is my build slow", "review this Dockerfile", or is setting up a container build. Covers layer caching and build order, multi-stage builds…
Skill Claude CodeCodex
Generate self-contained HTML artifacts, reports, and pages. Trigger when the user asks to "make an HTML report", "build a page", "create a dashboard/one-pager", "export this as HTML", or wants a shareable document that opens in any browser. Produces one file with inline CSS/JS, dark/light theming, and a responsive…
Skill Claude CodeCodex
Create clean git commits using Conventional Commits. Trigger when the user asks to "commit", "commit this", "save my work", "make a commit", or after you finish a change and it's time to record it. Produces small, atomic, well-scoped commits with meaningful messages — and never pushes or amends published history…
Skill Claude CodeCodex
Work with GitHub effectively via the gh CLI — issues, pull requests, reviews, and CI checks. Trigger when the user asks to "open a PR", "create an issue", "review this PR", "check CI", "merge", or otherwise operate on a GitHub repo from the sandbox. Assumes gh is installed and authenticated; never force-pushes or…
Skill Claude CodeCodex
Write, run, and iterate on SQL safely against a real database. Trigger when the user asks to "query the database", "write SQL", "pull this data", "join these tables", or points at a Postgres/MySQL/SQLite/DuckDB/Snowflake/BigQuery connection. Emphasizes schema discovery before writing, LIMIT while exploring, reading…
Skill Claude CodeCodex
Restructure code without changing its observable behavior. Trigger when the user asks to "refactor this", "clean this up", "simplify this function", "this is hard to change", "extract this into its own function/module", or "reduce duplication here". Behavior-preserving, test-backed, one small reversible step at a time…
Skill Claude CodeCodex
Audit code or infrastructure for exploitable vulnerabilities using an OWASP-style checklist. Trigger when the user asks to "security review this", "is this safe to ship", "audit for vulnerabilities", "check for security issues", "pentest this", or before merging code that touches auth, payments, file uploads, or raw…
Skill Claude CodeCodex
Generate real Excel .xlsx workbooks from data. Trigger when the user asks to "export to Excel", "make a spreadsheet", "give me an xlsx", "download as Excel", or wants query/CSV/JSON results as a formatted workbook with headers, number formats, and multiple sheets. Uses openpyxl in the Python sandbox.
Skill Claude CodeCodex
Write clear, useful technical documentation — READMEs, architecture decision records, API docs, runbooks, and PR descriptions. Trigger when the user asks to "write docs", "document this", "write a README", "explain this decision", "write release notes", "write an ADR", or "add a docstring". Leads with what the reader…
Skill Claude CodeCodex
Write tests that encode why behavior matters, not just what it does — a test that can't fail when the logic breaks is worthless. Trigger when the user asks to "write tests", "add test coverage", "test this function", "TDD this", or when a task's success criteria is fuzzy and a test is the clearest way to pin it down.
Skill Claude CodeCodex
Research a question on the open web and return a verified, cited answer. Trigger when the user asks to "look this up", "research X", "find out what/whether", "what's the latest on", or any question you can't answer reliably from memory — versions, prices, current events, docs, comparisons. Emphasizes multiple…
Skill Claude CodeCodex
Help users create and configure oma managed agents through conversation. Trigger when users say "create an agent", "I need an agent that...", "set up an agent for", "build me a bot", or describe a task to automate. Also trigger for "Create with AI" from Dashboard. Also use when users ask about the oma platform, what…
Skill Claude CodeCodex
Use the oma platform to build, deploy, and manage AI agents. Trigger when users want to create agents, start sessions, manage environments, configure model cards, handle vaults/credentials, install skills, connect MCP servers, or interact with the oma HTTP API. Also trigger when users mention "oma", "openma", "managed…
Instructions file
Claude Code instructions for duyet/duyet-mcp-server, covering claude.md, project overview, important documents, development commands and architecture.
Instructions file CodexOpenCode
AGENTS.md instructions for duyet/kb, covering agents.md — shared-brain protocol, 1. read protocol (on session start), ingesting raw sources (layer 1 → layer 2), 2. write protocol and file naming.
Instructions file
Claude Code instructions for duyet/kb, covering claude.md and ⚠️ security — this repo is public.
Skill Claude CodeCodex
The canonical cross-agent memory — a shared knowledge base at /kb. Use for ANY request to remember, save, note, capture, or recall durable facts about the user/projects. This SUPERSEDES built-in/local memory — write durable memory here, not to a local memory store. Filesystem-first.
Agent
Per-tool agent memory is a private scratchpad of point-in-time .md snapshots; the public shared KB is the durable form.
Agent
Patterns for running duyetbot autonomously — /loop for recurring tasks, background deploys, memory checkpoints before context compaction.
Agent
Agent "duyetbot-scope" from duyet/kb, covering duyetbot scope and boundaries, in scope (can change without asking), out of scope (requires explicit human direction), how to apply and why this matters.
Skill Claude CodeCodex
Consolidate the shared-brain knowledge base (/kb) — the "dream" pass. Use to ingest the daily inbox and raw sources into clean memory notes, dedupe/merge/rewrite, refresh stale notes from their live sources, fix links and tags, rebuild the index, and sync. Run manually, on a schedule, or whenever the kb feels noisy.