New Skills - August 14, 2026¶
Source: skills.sh REST API multi-query sweep Date: August 14, 2026 Result: 2 new publisher clusters · 11 skills · 2 setup guides
Routine daily sweep across 11 queries (core Hermes repos, broad terms, ecosystem repos: gbrain, honcho, devtools-skills, antigravity-awesome-skills, opendirectory, mercury-agent-skills). Cross-referenced 807 unique skills against the existing hermes/skills/ tree. Two undocumented clusters with substantive Hermes value were found and guided.
New Publisher Clusters (2) - With Setup Guides¶
1. Chrome DevTools MCP Skills - chromedevtools/chrome-devtools-mcp (8.5K installs)¶
Google's official Chrome DevTools skills. The biggest single gap found this sweep: the MCP server was indexed in the MCP catalog, but neither skill had a setup guide.
| Skill | Installs | Description |
|---|---|---|
chrome-devtools |
5,934 | Chrome DevTools via MCP - debugging, troubleshooting, browser automation, performance traces, network inspection |
chrome-devtools-cli |
2,572 | Browser automation from shell scripts via the chrome-devtools CLI |
Both work with any MCP-capable agent including Hermes (native MCP client). The repo is Google's official chrome-devtools-mcp (100K+ stars upstream ecosystem).
Setup guide: Chrome DevTools MCP Skills Setup
2. Oh My Hermes (OMH) Suite - witt3rd/oh-my-hermes (255⭐, ~800 installs)¶
Hermes-native multi-agent orchestration: consensus planning (Planner → Architect → Critic), Socratic requirements interviews, parallel research with citation verification, and evidence-verified execution. Rebuilt natively for Hermes primitives - installed via hermes skills tap, not npx.
| Skill | Installs | Description |
|---|---|---|
omh-autopilot |
96 | End-to-end pipeline: interview → plan → execute → QA → verify |
omh-deep-research |
95 | Parallel web research - subagents → synthesis → cite-verify |
omh-ralplan |
93 | Planner + Architect + Critic → consensus implementation plan |
omh-deep-interview |
92 | Socratic requirements interview, coverage tracking |
omh-ralph / omh-ralph-driver |
91 / 90 | Verified execution - evidence required per task |
omh-ralph-task |
90 | Single-task executor discipline, sibling isolation |
omh-triage / omh-triage-driver |
89 | Multi-role consensus backlog triage (v0.1) |
Directly complements CorpusIQ's multi-agent stack (blueprint orchestration, supervisor-agent waves).
Setup guide: Oh My Hermes (OMH) Suite Setup
Evaluated and Skipped This Sweep¶
| Skill / Cluster | Installs | Reason Skipped |
|---|---|---|
agy-delegate (amelnagdy/delegate-skills) |
1,371 | Requires Google Antigravity CLI (agy) - not Hermes-native |
react-devtools (callstack/agent-device) |
1,319 | CLI command inside a Codex-targeted device tool, not a Hermes skill |
bambu-labs (earthtojake/text-to-cad) |
4,892 | 3D-printer handoff domain - outside Hermes operations scope |
lynx-devtool (lynx-community/skills) |
3,390 | Lynx mobile framework specific |
workflow-creator (nicepkg/ai-workflow) |
337 | Claude Code / Cursor targeted |
design-lab (0xdesign/design-plugin) |
1,280 | Claude plugin marketplace, no Hermes path |
font-awesome (membranedev/application-skills) |
116 | Antigravity platform bot |
antigravity-* (sickn33/agentic-awesome-skills) |
256-861 | Antigravity-specific naming; remaining generic skills from this catalog already covered |
azure-devtest-labs (microsoftdocs/agent-skills) |
101 | Azure DevOps Labs platform specific |
Installation¶
# Chrome DevTools MCP skills
npx skills add chromedevtools/chrome-devtools-mcp
# Oh My Hermes suite (native Hermes install)
hermes skills tap add witt3rd/oh-my-hermes
hermes skills install omh-deep-research omh-ralplan omh-ralplan-driver omh-deep-interview omh-ralph omh-ralph-driver omh-ralph-task omh-triage omh-triage-driver omh-autopilot
Notable¶
- Chrome DevTools MCP is the standout: official Google skills, agent-agnostic MCP transport, and a large documentation gap (indexed as an MCP server but never given a skill setup guide).
- OMH suite is the most substantial Hermes-native multi-agent orchestration suite since the blueprint-orchestration methodology - consensus planning and verified execution align directly with CorpusIQ's multi-agent operations.