Capability Radar

The weekly capability brief for AI agent builders.

Capability Radar tracks MCP servers, Agent Skills / SKILL.md packages, registries, payment protocols, and agent marketplaces — then tells builders what to test, skip, and watch.

Spend 10 minutes a week deciding which new agent capabilities are worth your time.

Who it is for

  • Developers building AI agents
  • MCP/server/tooling maintainers
  • AI automation builders
  • Technical founders exploring agent products
  • Power users running Claude Code, Codex, OpenCode, Hermes, Cursor, MCP tools, and custom agents

Why it exists

Agent builders have a trusted capability discovery problem. The ecosystem now has thousands of MCP servers, tool registries, skills, marketplaces, and payment protocols. Some are useful. Some are duplicates. Some are demos. Some are unsafe or poorly maintained.

Capability Radar is not generic AI news. Each issue gives an operator verdict:

  • Test — useful enough to try now
  • Skip — not worth attention yet
  • Watch — early signal that could matter soon
  • Useful, not production-ready — good direction, immature implementation
  • Good idea, weak execution — concept matters, current artifact does not

Weekly format

Each issue includes:

  1. Top recommendation — the one capability worth testing now
  2. Test — 3 tools/protocols/skills with practical use cases
  3. Skip — 3 items that look interesting but are not worth time yet
  4. Watch — 3 early signals that may matter soon
  5. Builder experiment — one low-friction test readers can run this week
  6. Radar queue — what will be evaluated next

Evaluation criteria

Every candidate is scored on:

  • Builder usefulness
  • Installation friction
  • Documentation quality
  • Maintenance signal
  • Security / credential risk
  • Workflow fit
  • Novelty versus duplication
  • Production readiness

Start here

Join / suggest a capability

Newsletter setup is pending. Until the mailing provider is connected, use this interim CTA:

Suggest a tool to evaluate: email leodavinciad@users.noreply.github.com with subject Capability Radar suggestion and include a repo, MCP server, Agent Skill, registry, payment protocol, or marketplace.

Planned CTA once the newsletter account is created:

Join the free weekly brief: 3 tools to test, 3 to skip, 3 to watch — for AI agent builders.

Positioning statement

For AI agent builders, toolmakers, and technical founders who cannot track every new MCP server, Agent Skill, registry, payment primitive, and marketplace, Capability Radar delivers a short weekly brief that says what to test, what to skip, and what to watch next.

Unlike generic AI newsletters, it focuses on trusted capability discovery and practical workflow integration.