Integration proposal: Hermes Agent (auto-drafted 2026-06-29)

[NativeTeachingAidsB]

Score: 13 — highest-signal candidate this week. Labels to add manually: integration-proposal, auto-generated, urgent.

Auto-generated by the weekly-landscape-scan routine. Draft spec in docs/superpowers/specs/2026-06-29-integrate-hermes-agent.md (on branch docs/landscape-report-2026-06-29). Do not promote to approved without human review.

What is it?

Hermes Agent is an open-source autonomous AI agent by Nous Research — currently the most-starred general-purpose AI agent on GitHub at 205,000+ stars, 37,000+ forks. v0.17.0 released June 19, 2026. Ships full MCP integration, Skills Hub marketplace, parallel subagent spawning, and TTS via OpenAI Tool Gateway.

Why it matters for codetalker

Hermes already has basic TTS (OpenAI API, requires key). Codetalker adds: local Piper voices (no API key), coding-specific cadence modes, 3D animated character avatars. Additive, not competing.

Integration path is identical to Cursor/Cline: add codetalker-mcp to ~/.hermes/config.yaml. No codetalker code changes needed.

Stretch goal: publish codetalker as a Hermes community skill on agentskills.io.

Scoring

Signal	Score
Active (v0.17.0, June 19)	+2
>500 stars	+1
>2000 stars	+2
MCP support (first-class)	+2
Skills Hub (plugin/skill system)	+1
Free tier (MIT)	+1
Mentioned in 2+ sources	+2
HN/PH coverage	+2
Total	13

Open questions (need human decisions)

1. Position codetalker as TTS replacement (local Piper, no API key) or purely additive (character avatars + coding cadence)? Affects hermes.md copy.
2. Skills Hub (agentskills.io) submission — self-serve or approval required?
3. Does the codetalker daemon handle concurrent MCP connections from Hermes subagents?

Effort

Small — 1–2 days. Integration guide only; MCP wiring is standard.

---

Routine run: 2026-06-29 | Landscape report: docs/landscape-reports/2026-06-29.md