Browse Skills

Checkpoint saves a structured state snapshot before stopping or handing off, capturing the active plan, recent decisions, file pointers with line numbers, open questions, and next actions. Use it before a break, model switch, or collaborator handoff when you need to resume quickly.

Compress Session distills the current Claude Code conversation into a structured note (decisions made, open questions, file pointers, next 1–3 actions) saved to quality_reports/session_logs/ before auto-compaction truncates context. Use when context is approaching the auto-compact threshold, after long pipelines, or when the user asks to compress or distill the session.

Context Status displays the current session state, including context usage estimate, active plan, session log, and preservation state. Use it to check session health and whether auto-compact is approaching.

Learn extracts reusable knowledge from the current session into a persistent skill. Use it when you discover something non-obvious, create a workaround, or develop a multi-step workflow that future sessions would benefit from.

New Skill scaffolds a new Claude Code skill by interviewing the user for purpose, trigger phrases, arguments, and tool needs, then writes a convention-compliant SKILL.md that passes integrity gates on first try. Use when creating a new slash command or turning a repeated workflow into a skill.

Research Interview asks clarifying questions one at a time to turn a fuzzy research idea into a structured spec covering research question, hypotheses, identification, data needs, and empirical strategy. Use it when you say "interview me" or need help formalizing a half-baked research idea into a project.

Agent Optimization improves one Test Agent through an evidence → hypothesis → Candidate → evaluation → accept or rollback loop. It uses a frozen Benchmark, Scoreboard, and public Test Traces as black-box feedback while delegating all evaluation to an agent-evaluation subagent.

AgentHub Models calls model APIs through @prismshadow/agenthub for streaming text, image generation, speech synthesis, embeddings, and the supported-model registry with one client. Use it when building AI apps that need a single TypeScript client for multiple model providers.

Benchmark Design builds and calibrates a multi-Case capability benchmark for one Test Agent and establishes a traceable Formal Baseline. Use it when you need to design, pilot, and freeze a benchmark without changing the test agent.

LlamaFactory fine-tunes open-weight LLMs (LoRA/QLoRA, full-parameter, SFT, DPO) via YAML configs and the llamafactory-cli command. Use it to register datasets, train models, merge LoRA adapters, and serve or chat with the result.

Ollama Serving pulls and runs local models with Ollama, exposing an OpenAI-compatible endpoint on localhost:11434 for apps and agents. Use it to add a self-hosted model to PenguinHarness or any OpenAI client.

Penguin CLI manages model API keys, default models, and per-agent vault secrets for the Penguin harness. Use it when configuring model credentials and agent secrets.