hermes-CCC

Compress conversation context — summarize the current session, extract key decisions and facts, then compact history to free up context window.

Generate insights about your Claude Code usage — what topics you work on most, common patterns, productivity trends.

Route Claude Code work by complexity, risk, and tool needs. Use when deciding how much reasoning depth a task needs, whether to read project memory first, whether the task should be decomposed, and whether the work is lightweight, standard, or investigation-heavy.

Query Polymarket prediction markets for probability data and research insights on real-world events.

Search and retrieve academic papers from arXiv using their free REST API. No API key needed.

Query Base (Ethereum L2) blockchain data — wallet balances, token info, transactions, gas analysis, contract inspection. No API key required.

Monitor and summarize blog posts, RSS feeds, and web content for research and staying current with topics.

Open-source embedding database for RAG — store embeddings, vector search, metadata filtering. Simple API, scales from notebook to production.

Manage Docker containers, images, volumes, networks, and Compose stacks - lifecycle ops, debugging, cleanup, and Dockerfile optimization.

Free web search via DuckDuckGo - no API key needed. Text, news, images. Use as fallback when other search tools unavailable.

Create and export diagrams with Excalidraw - whiteboard-style sketches, flowcharts, architecture diagrams via MCP or CLI.

Facebook AI Similarity Search — ultra-fast vector similarity search for large-scale local embedding retrieval without a database server.

Optimize transformer attention with Flash Attention — 2-4x speedup, 10-20x memory reduction for long sequences on CUDA GPUs.

Review GitHub pull requests with a findings-first engineering mindset. Use when auditing diffs for bugs, regressions, security issues, missing tests, or risky design choices, and when producing actionable review comments instead of generic summaries.

GitHub issue management — create, update, label, milestone, and link issues to PRs via gh CLI or curl fallback.

Run a disciplined GitHub pull request workflow from branch creation through merge. Use when opening a PR, updating a PR after review, preparing review-ready diffs, managing draft versus ready states, or landing changes with clear validation and risk notes.

Automate Google Workspace — Gmail, Drive, Sheets, Docs, Calendar via Google API Python client or gcloud CLI.

Expert guidance for GRPO/RL fine-tuning with TRL for reasoning and task-specific model training.

Manage durable project memory for Claude Code. Use when prefetching relevant history before work, syncing stable decisions after work, nudging for unsaved lessons, compressing noisy notes, or auditing memory quality.

Switch Claude's personality/persona between specialized modes — researcher, coder, analyst, creative, advisor.

Search across past conversation sessions, memory files, and project history to recall what was discussed, decided, or built in previous sessions.

Create, improve, inventory, and audit Claude Code skills. Use when turning a solved workflow into a reusable skill, expanding an existing skill, checking trigger quality, or standardizing a skill pack.

Capture Claude Code interaction trajectories in training-friendly formats. Use when saving successful runs, failed runs, review samples, evaluation traces, or compact datasets for later QA and fine-tuning analysis.

AI-native cross-session user modeling — builds a persistent, growing model of who you are, your expertise, preferences, and working style across all Claude sessions.

HuggingFace Hub — download models/datasets, upload artifacts, search, and manage tokens via CLI and Python API.

Structured LLM outputs with Instructor — Pydantic models as response schemas for OpenAI, Anthropic, and any OpenAI-compatible API.

Jupyter notebook and kernel operations — start kernels, execute cells programmatically, export, and data analysis patterns.

Linear project management — create issues, manage cycles, update projects, and query team progress via Linear API or MCP.

Run quantized LLMs locally with llama.cpp — CPU+GPU inference, GGUF format, OpenAI-compatible server, and Python bindings.

Create mathematical animations and explainer videos with Manim Community - code-driven animations for math, CS concepts, data visualization.

Convert any CLI tool or Python function into an MCP server using fastmcp. Zero-config tool injection into Claude Code.

MCP server integration in Claude Code — add servers to .mcp.json, configure transports, and use their tools natively.

1Password CLI integration — retrieve secrets, manage vaults, inject credentials into scripts without hardcoding.

Open-source security forensics - analyze repositories, dependencies, and code for malicious patterns, supply chain risks, and vulnerabilities.

Managed vector database for production RAG — serverless and pod-based deployment, hybrid search, namespaces, and metadata filtering.

High-performance vector search engine for production RAG — Rust-powered, horizontal scaling, hybrid dense+sparse search, metadata filtering.

Academic paper writing workflow — literature review, LaTeX setup, structure, citations, and conference submissions.

OSINT username search across 400+ social networks using the Sherlock Project.

Query Solana blockchain data — SOL balances, SPL tokens, transactions, programs, and NFTs via JSON RPC API.

Run Stable Diffusion locally with diffusers — text-to-image, img2img, inpainting, ControlNet, and SDXL.