hermes-agent-evolution

Configure, extend, or contribute to Hermes Agent.

Merge ready, green-CI evolution PRs into main and self-update

Write desktop app plugins that add UI panes and commands.

Drive the user's desktop in the background — clicking, typing, scrolling, dragging — without stealing the cursor, keyboard focus, or switching virtual desktops / Spaces. Cross-platform: macOS, Windows, Linux. Works with any tool-capable model. Load this skill whenever the `computer_use` tool is available.

Analyze the agent's real sessions with users to find what blocks practical task completion, and turn those findings into improvement issues

Audit memory health: secrets, stale, broken wikilinks.

Use when the user wants to do anything in Unreal Engine through Epic's official editor-embedded MCP server (catalog entry: unreal-engine) — build/light/populate scenes, place and transform actors, author Blueprints, animate with Sequencer, create material instances, frame cameras, take screenshots, render, import assets, run PIE test sessions and automation tests, or automate the editor end-to-end from plain-English prompts with no Unreal knowledge required. Covers the tool-search discovery walk (list_toolsets/describe_toolset/call_tool), serial game-thread call discipline, ProgrammaticToolset batching, the Blueprint graph DSL loop, scene-craft numbers (physical light units, exposure, scale conventions), complete build recipes, save/undo hygiene, and extending the tool surface with custom Python toolsets.

Manual OAuth for remote MCP servers on headless gateways.

Agent2Agent (A2A) interoperability for Hermes. Slice 1 exposes Hermes' existing tools and skills as an A2A Agent Card served at /.well-known/agent.json, so other A2A-speaking agents can discover what this Hermes instance can do. Discovery only -- no A2A JSON-RPC server yet.

Analyze issues and PRs to prioritize implementation

Implement selected issues and self-update

Keep the fork current with upstream Hermes Agent by merging its published RELEASE tags (not bleeding-edge main)

Research other AI agents, papers, and trends for Hermes Evolution improvements

Extract the concrete technique from ONE high-value research paper into a structured, validated draft the pipeline can act on

Create GitHub issues and PRs based on research findings

Join a Google Meet call, transcribe live captions, optionally speak in realtime, and do the followup work afterwards. Use when the user asks the agent to sit in on a meeting, take notes, summarize, respond in-call, or action items from it.

Plan, set up, and monitor a multi-agent video production pipeline backed by Hermes Kanban. Use when the user wants to make ANY video — narrative film, product/marketing, music video, explainer, ASCII/terminal art, abstract/generative loop, comic, 3D, real-time/installation — and the work warrants decomposition into specialized profiles (writer, designer, animator, renderer, voice, editor, etc.) coordinated through a kanban board. Performs adaptive discovery to scope the brief, designs an appropriate team for the requested style, generates the setup script that creates Hermes profiles + initial kanban task, then helps monitor execution and intervene when tasks stall or fail. Routes scenes to whichever Hermes rendering / audio / design skill fits each beat (`ascii-video`, `manim-video`, `p5js`, `comfyui`, `touchdesigner-mcp`, `blender-mcp`, `pixel-art`, `baoyu-comic`, `claude-design`, `excalidraw`, `songsee`, `heartmula`, …) plus external APIs for TTS, image-gen, and image-to-video as needed.

Migrate a user's OpenClaw customization footprint into Hermes Agent. Imports Hermes-compatible memories, SOUL.md, command allowlists, user skills, and selected workspace assets from ~/.openclaw, then reports exactly what could not be migrated and why.

Audit a completed task against the user's original request before declaring done. Catches omitted constraints, misread scope, partially-met requirements, and unsupported "it's done" claims after long multi-tool loops. Opt-in quality gate — run it, report the verdict, never silently re-loop.

Decompose a research sub-task into N worker prompts, fan them out via delegate_task, collect candidate outputs

Use when building, modifying, reviewing, or auditing an AI agent / agentic or LLM system, or whenever an agent processes external or untrusted data (email, documents, web pages, RAG context, tool or MCP descriptions, API responses). Triggers include agent security audit, prompt injection, indirect prompt injection, tool poisoning, jailbreak, RAG or vector-DB security, MCP server risks, unsafe LLM output handling, over-privileged agent tools, sensitive data or PII leakage, Denial of Wallet.