frumu-aiOfficial·98 Agent Skills Included

tandem

Runtime policy enforcement, approvals, and audit for agent workflows

Enforces runtime policy between agents and the tools, data, memory, and actions they use. Blocks unauthorized tool calls, pauses consequential actions for human approval, and scopes memory access per tenant. Records artifacts, decisions, and audit evidence outside the model so teams can run governed automations safely.

114|13

↗

npx skills add frumu-ai/tandem --all -g -y

All Skills in This Repository (98)

Pure Emerald Level Indicators

📦 In Repo

frumu-ai

social-media-scheduler

Draft and schedule social media posts for specified channels and themes.

news-digest

Gather and summarize topic-specific news articles on a recurring schedule.

competitor-price-tracker

Monitor competitor product pages for price changes and send alerts.

email-digest

Generates a daily morning digest of summarized, unread emails from your inbox.

github-issue-reporter

Convert production incident summaries into structured GitHub issues with deduplication.

research-agent

Research specified topics and synthesize findings into cited briefs.

dev-agent

Automates the full MLDSP pipeline for bug-fix and feature-request TDD in JavaScript/TypeScript repos.

website-change-monitor

Detect significant content changes on specified web pages and alert.

server-monitor

Monitor service endpoint health and latency, aggregating failures to trigger incident alerts.

notion-weekly-report

Generate weekly progress summaries from Notion sources and email them to a recipient.

security-playbook-builder

Generate prioritized security controls, response runbooks, and compliance documentation.

micro-drama-script-studio

Generate episode scripts, beat outlines, cast sheets, and writer's room dashboards.

Official

Intermediate

Frequently Asked Questions

FAQPage Schema

How to install Tandem?▼

Run `npx skills add frumu-ai/tandem --all -g -y` in your terminal to install all skills in this collection globally.

What does Tandem do for AI agents?▼

Tandem acts as a runtime control plane that decides which tools agents can call, which actions need human approval, and what memory they can access. It records every decision and artifact in an audit trail outside the model.

What skills are included in Tandem?▼

The collection includes templates for marketing, sales, legal review, finance, data analysis, support triage, research, and bioinformatics, plus mission blueprint compilers for building staged multi-step workflows.

Can non-developers use Tandem skills?▼

Yes. Many skills like email digests, sales briefings, and content planning work from plain-English requests, while the runtime handles permissions and approvals automatically.

Does Tandem work with different model providers?▼

Yes. The runtime is provider agnostic and works with OpenRouter, Anthropic, OpenAI, or local Ollama endpoints.