Skill Claude CodeCodex
Inspect a MoneyClaw wallet, create bounded payment tasks, and continue user-confirmed payment steps using a prepaid account. Use only when the user clearly asks to use MoneyClaw for their own payments.
Plugin Claude Code
Plugin marketplace listing 2 plugins: onlyharness, superskill.
Instructions file CodexOpenCode
AGENTS.md instructions for elvismusli/onlyharness, covering superskill agent guide, what this repo is, dev commands, agent use and superskill local/internal-alpha contract.
Instructions file
Claude Code instructions for elvismusli/onlyharness: Read AGENTS.md first. Use the documented local checks before proposing a commit. Never commit infra/production.env, tokens, cookies, or service-role keys. Production deploys use the repo's in-place deploy scripts and must be explicitly requested. For harness work…
Agent
Summarize adjacent agent harness projects and keep discovery separate from installation.
Plugin Claude Code
Find, inspect, install and publish reusable AI-agent resources and native harnesses from onlyharness.com.
MCP server Claude CodeCodexCursor +2
MCP server "onlyharness", hosted remotely at onlyharness.com, as configured in elvismusli/onlyharness.
Skill Claude CodeCodex
Use when the user wants a ready-made AI-agent resource or native harness package (research, support triage, finance safety review, GTM): search, suggest, pull, run, eval, gate, update, or publish one from onlyharness.com.
Plugin Claude Code
Install once from a pinned link, then find and consent to an exact reviewed capability.
Skill Claude CodeCodex
Use when the user explicitly asks SuperSkill to find a reviewed capability, search the public skill catalog, or install an explicitly selected SuperSkill-hosted skill; also use when a substantial task maps to a curated job category and no selected local skill covers it.
Agent
Create minimal eval cases that catch behavior regressions.
Agent
Map source workflow to harness.yaml without inventing missing details.
Agent
Extract stages, roles, inputs, outputs and caveats from source markdown.
Agent
Check freshness, sample size, joins, denominator drift and conflicting cuts.
Agent
Write a go/no-go data quality verdict for the business decision.
Agent
Map every metric to source tables, event names, filters, windows and owners.
Agent
Find weak assumptions, contradictions, missing sources and overclaiming.
Agent
Turn research into a structured answer with clear evidence and gaps.
Agent
Collect source-backed market evidence and name confidence limits.
Agent
Extract ledger, provider, user-facing and retry paths.
Agent
Write required verification gates before merge or release.
Agent
Flag double charge, double credit, withdrawal and provider mismatch risks.
Agent
Extract decision, constraints and real options.
Agent
Lead with recommendation and keep caveats explicit.