skills

LSP-enable a Python, TypeScript, or Rust repository for fast-agent development. Use when creating or refreshing agent cards with LSP function tools, configuring ty, typescript-language-server, or rust-analyzer integration, or scoping repo-local code navigation safely.

Guide for implementing fast-agent hooks and user-invoked command actions. Use when adding hook functions to agent cards or Python code for tasks like history compaction, saving sessions, modifying tool calls, managing agent lifecycle, or adding explicit slash-command add-ins with PluginCommandActionContext.

Create polished standalone single-file HTML engineering artifacts in the Birchline design language. Use for session summaries, module/process explainers, implementation plans, design direction comparisons, PR/change writeups, findings-first code reviews, incident/status reports, slide decks, flow diagrams, and compact numeric/data briefs where the output should be a beautiful local HTML document with embedded CSS and no external assets.

Build, review, or refactor GEPA optimization loops that use fast-agent. Use when Codex needs to design GEPA evaluators, Actionable Side Information, Pareto/frontier metrics, FastAgentReflectionLM, FastAgentBatchEvaluator, BatchRunner, EvalRun/CandidateRun artifact layouts, Trackio monitoring, or refactor OpenClaw/Birch-style prompt, policy, skill, batch, or artifact-generation optimizers.

Investigate fast-agent session and history files to diagnose issues. Use when a session ended unexpectedly, when debugging tool loops, when correlating sub-agent traces with main sessions, or when analyzing conversation flow and timing. Covers session.json metadata, history JSON format, message structure, tool call/result correlation, and common failure patterns.

Automate fast-agent runs from CLI, Docker, and Hugging Face Jobs. Use when users need repeatable non-interactive execution (`fast-agent go --message` / `--prompt-file`), deterministic artifact handling (`--results`), multi-model fan-out, containerization with uv, or scheduled/cloud execution with hf jobs and secrets management.

Deploy general static websites and frontend apps to Hugging Face Spaces using `sdk: static`. Use when the user wants to host HTML/CSS/JS sites or built SPA frontends (React, Vite, Svelte, Vue, etc.) on Spaces for convenient sharing. Include setup for README YAML metadata (`app_file`, optional `app_build_command`), deployment via `hf` CLI or Python API, GitHub Actions sync, and static-Space troubleshooting.

Implement hook-based history compaction in fast-agent (rolling window, context threshold truncation, tool-result clearing, summary prompt compaction). Use when adding or extending compaction hooks and wiring them into agent cards.

Deploy fast-agent MCP Servers to Hugging Face Spaces using Docker. Use when the user wants to deploy an "agent card" as an MCP Server. Includes templates and CLI commands for deploying agent cards with custom tools to HF Spaces.