ProfSynapse

Read, search, and edit the user's Obsidian vault (notes, folders, canvas, tasks, memory/workspaces, saved prompts) from the shell via the `nexus` CLI — no MCP connection needed. Use whenever the user refers to their vault, notes, daily notes, second brain, or Obsidian, or asks you to find/read/change something stored there.

Grade LLM models on Nexus tool-use (the two-tool getTools/useTools protocol) with the live eval harness in tests/eval/. Use when asked to grade, benchmark, or evaluate one or more OpenRouter (or other provider) models for how well they drive Nexus tools — e.g. "grade google/gemma-4-31b-it" or "benchmark these models on our harness".

Version bump and GitHub Actions release for the Nexus Obsidian plugin. Use when the user wants to cut a release, bump the version, or publish a new version after stable changes are ready.

Add, update, or verify Nexus LLM provider model definitions. Use when adding newly released models, changing OpenAI/OpenRouter/Codex/GitHub Copilot/Anthropic/Google model metadata, updating provider defaults, or live-testing whether a model ID works through the reusable provider smoke test.

Version bump and GitHub Actions release for the Nexus Obsidian plugin. Use when the user wants to cut a release, bump the version, or publish a new version after stable changes are ready.

Version bump and GitHub Actions release for the Nexus Obsidian plugin. Use when the user wants to cut a release, bump the version, or publish a new version after stable changes are ready.

Run or update the Nexus LLM eval harness for arbitrary provider/model matrices. Use when comparing models in the real vault-like tool environment, changing eval configs, adding eval scenarios, or debugging eval reports.

Add, update, or verify Nexus LLM provider model definitions. Use when adding newly released models, changing OpenAI/OpenRouter/Codex/GitHub Copilot/Anthropic/Google model metadata, updating provider defaults, or live-testing whether a model ID works through the reusable provider smoke test.

Complete reference for the fine-tuning pipeline (SFT, KTO, GRPO), cloud HF Jobs workflows, autonomous experiment search, checkpoint evaluation, and LoRA surgery. Covers training CLI flags, YAML configuration, model presets, dataset requirements, LoRA settings, training monitoring, hyperparameter search, and post-training optimization. Use when training models, configuring training runs, choosing hyperparameters, running cloud experiments, inspecting HF jobs, or troubleshooting training issues. This skill is about USING the training system via CLI and YAML — never modifying source code.

Turn local agent transcripts (Claude Code, Codex CLI, or any format with an adapter) into high-quality fine-tuning rows. Use when someone wants to mine their own coding-agent conversation logs into an SFT/KTO/DPO/GRPO dataset — parse the logs, extract substantive turns, scrub secrets, score each session by outcome (tests passed / built clean / committed / PR'd), dedup, and tier by quality. Config-driven and format-pluggable; point it at a transcript directory and it produces deduped, quality-tiered JSONL. This skill is about USING the checked-in distill engine via CLI + YAML.

Complete reference for model upload and deployment. Covers HuggingFace upload, save strategies (LoRA, merged 16-bit, merged 4-bit), GGUF conversion, model merging, model cards, and the full upload workflow. Use when uploading models, creating GGUF files, merging LoRA adapters, or deploying to HuggingFace. This skill is about USING the upload/deployment tools via CLI — never modifying source code.

Complete reference for the SynthChat synthetic dataset generation system. Covers CLI commands (generate, improve, validate), scenario YAML authoring, rubric YAML authoring, settings configuration, evaluation, and full workflow. Use when generating datasets, writing rubrics/scenarios, configuring models/workers, improving dataset quality, or running evaluations. This skill is about USING the system via CLI and YAML — never modifying source code.

Complete reference for the embedding fine-tuning pipeline (SentenceTransformers bi-encoders) — the `embedding` training method, model registry, dual loader (Unsloth fast path + ST fallback), adapter modes (full/lora/frozen_head), triplet/pairs dataset format, retrieval evaluation (recall@k/MRR/nDCG/MAP via the `retrieval` verifier), and the local-Docker smoke recipe. Use when training a retrieval embedder, picking a base model + adapter mode, preparing triplet data, or evaluating a fine-tuned embedder against a labeled corpus. This skill is about USING the embedding pipeline via CLI and YAML — never modifying source code.

Complete reference for the config-first model evaluation system. Covers the Evaluator CLI, assertion-driven YAML scenarios, response views, backend configuration, presets, scoring, LLM-as-judge, model comparison, and HuggingFace integration. Use when evaluating models, writing test prompts, comparing training runs, or interpreting eval results. This skill is about USING the evaluation system via CLI and YAML.

Publish local dataset artifacts to a Hugging Face dataset repo. Use when uploading a JSONL dataset, pushing a filtered dataset variant, syncing a matching .metadata.json sidecar, or renaming a dataset file in the target repo. This skill is about USING the checked-in dataset publish script via CLI — never ad hoc Python.

End-to-end case studies showing how to implement the full training pipeline for different skill types. Covers three complete worked examples — tool-calling training, essay-style training, and agentic search (RAG agent) training — demonstrating dataset design, synthetic generation, validation, fine-tuning, evaluation, and iteration. Use when onboarding to the project, understanding how all components fit together, explaining the pipeline to others, or planning a new training capability. This skill is about UNDERSTANDING the system holistically — reference the other skills for specific CLI commands.

Author, organize, and launch tuner-backed mechanistic-interpretation cells (activation reading and writing) for the Epistemic-Humility project. Use when designing a new steering / extraction / probe-fit / gate-scoring cell, deciding where its config, direction, grader, and outputs live, or launching one locally or on Modal. This skill is about USING the synaptic-tuner `mechinterp` verbs via declarative recipe YAML plus project plug-ins; it never modifies the tuner submodule and never touches the frozen legacy machinery.

Operational runbook for the Epistemic-Humility Locked Matrix Experiment Runner - expands the PROTOCOL v0.3 (LOCKED) run matrix (3-seed headline + LR/beta sensitivity panel at 4B, 3-seed confirm at 8B, 2 bridge replication cells) into per-cell tuner invocations across two lanes (local RTX 3090 / HF Jobs cloud), with hard pre-registration count assertions, prerequisite gating, data staging, and committed provenance run records. Use when launching, dry-running, or gating the locked matrix, materializing per-cell recipes, or inspecting run records. This skill is about USING the runner via checked-in scripts; it never modifies the synaptic-tuner submodule.

Lifecycle tooling for the experiments-first repo layout - one self-contained directory per evidence-producing experiment (steer cell, training run, eval, probe-fit, or lab diagnostic) under experiments/, each with a signed AMENDMENT.md, a thin machine-readable experiment.yaml manifest, pinned instrument configs, and generated indices. Use to scaffold, sign, list, show, resolve, and validate experiments, to regenerate the registry, and to understand the draft -> sign -> run -> resolve lifecycle and the shared-input promotion rule. The bin/exp wrapper runs these scripts; validation and registry regeneration run at commit time via .githooks/pre-commit.

Package a terminal experiment's data exhaust (aggregate stats and, where license-permitted, row-level generation text) into reproducible Hugging Face datasets, with a license gate, containment lint, and a thin uploader. Use when an experiment under experiments/<slug>/ reaches a terminal verdict (resolved, null-result, or falsified) and its analysis-committed/ artifacts should be published.

Standardize the per-model layer-mapping extraction (full-depth hidden-state capture, workspace-dimension profile, three-axis held-out read panel) so any new model, family, or size entering the Epistemic-Humility program gets a fast, cheap, uniform layer atlas before any actuation design. Use when a new model/family/size enters the program, before designing any per-family actuation cell, or when an existing family's substrate changes (new revision, new checkpoint).

Create, validate, normalize, export, and analyze typed Obsidian knowledge graph metadata in the Professor Synapse vault, including `kg` node metadata, `relationships` edge objects, `related` link projections, ontology-governed edge types, legacy relationship shorthand, and graph analysis reports.

Run, plan, validate, or aggregate Epistemic-Humility local mechanistic-interpretability sweeps, including hidden-state candidate inventories, causal-pilot sweep planning, explicit non-GPU/GPU gates, base-original skip handling, and offline result aggregation. Use when working on local mech-interp sweeps, causal-pilot diagnostics, activation-addition/logit-diagnostic runs, behavior-axis scans, SAE feature screens, or future reruns of the legacy full candidate inventory.

Ingest a research paper (or a batch) into the library's atomic knowledge graph the Agents-K1 way. Use when you are handed a paper (URL, arXiv id, or a note already in library/) and you want it folded in as typed concept atoms, mechanisms, lineage edges, and a patched paper note, reconciled against the atoms already in the vault rather than re-extracted from scratch. Also use to backfill spine papers not yet atomized. This is the standing ingestion posture for growing the graph one paper at a time.

CRUA workflow for editing Office and proprietary documents (docx, xlsx, pptx, pdf) safely. Use when asked to modify, update, fill in, or rewrite a document file — checkout converts it to markdown/csv in _workspace/, you edit the copy, publish archives the old version and replaces the original. Never edit proprietary formats in place.

Handling Google Docs/Sheets/Slides files that appear in a Google Drive for desktop synced folder as .gdoc/.gsheet/.gslides files. Use when a task involves one of these extensions or when the user asks to edit a Google Doc from their local Drive folder.

Recovering lost or overwritten files from the guardrails archive store. Use when the user says a file was deleted, overwritten, truncated, or "messed up", or asks to roll back to an earlier version or undo a recent file operation.

Safety rules for working inside cloud-synced folders — OneDrive, SharePoint synced libraries, Google Drive for desktop, Dropbox. Use whenever a task touches files under a synced folder, before bulk operations, or when a file read returns empty/garbage content for a file that clearly should have data.

Use when user needs expert help, wants to summon a specialist, says "help me with", "I need guidance", or has a task requiring domain expertise. Creates and manages a growing collection of expert agents.

Use when user needs expert help, wants to summon a specialist, says "help me with", "I need guidance", or has a task requiring domain expertise. Creates and manages a growing collection of expert agents.

Use when user needs expert help, wants to summon a specialist, says "help me with", "I need guidance", or has a task requiring domain expertise. Creates and manages a growing collection of expert agents.

Build, improve, modularize, and validate Claude Code skills the right way. Use when creating a new skill, improving or auditing an existing one, refactoring a SKILL.md that grew too long, or adding validation scripts. Encodes a slim-router, progressive-disclosure, PACT-driven approach and is its own reference implementation.

Manages the Vapi voice AI platform API. Use when the user wants to create, list, update, or delete Vapi resources (assistants, calls, phone numbers, squads, tools, files, knowledge bases, blocks, sessions, campaigns, logs, evals, eval runs), make outbound calls, configure voice pipelines, run eval tests against assistants, or manage their Vapi account. Handles all CRUD operations against the Vapi REST API.