autoskill skill

by K-Dense-AI·K-Dense-AI/scientific-agent-skills·47k stars·MIT

Observe the user's screen via screenpipe, detect repeated research workflows, match them against existing scientific-agent-skills, and draft new skills (or composition recipes that chain existing ones) for the patterns not yet covered. Use when the user asks to analyze their recent work and propose skills based on what they actually do. Requires the screenpipe daemon (https://github.com/screenpipe/screenpipe) running locally on port 3030 — the skill has no other data source and will refuse to run if screenpipe is unreachable. All detection runs locally; only redacted cluster summaries reach the LLM.

A100/100content scan

Is the autoskill skill safe?

Clean: nothing in its files matched our rules. We read 14 files in the folder on 2026-09-28.

No findings.

Install the autoskill skill

A skill is a folder. Copy it into your agent's skills folder and the agent loads it when the task matches its description.

git clone --depth 1 https://github.com/K-Dense-AI/scientific-agent-skills.git /tmp/scientific-agent-skills
mkdir -p ~/.claude/skills
cp -r /tmp/scientific-agent-skills/skills/autoskill ~/.claude/skills/autoskill

available in every project

In the Claude apps, zip the folder and upload it from the Skills settings. The folder on GitHub

The instructions your agent would load

SKILL.md as published, without the frontmatter. Read it on GitHub

autoskill

Requires a running screenpipe daemon. This skill has no alternate data source — it reads exclusively from the local screenpipe HTTP API (default http://localhost:3030). If the daemon isn't running, run() raises ScreenpipeUnreachable with install instructions.

Network access & environment variables. This skill makes authenticated HTTP requests to (a) the user's local screenpipe daemon on loopback, and (b) the user-configured LLM backend — one of http://localhost:1234/v1 (LM Studio, default), https://api.anthropic.com (opt-in Claude), or a user-supplied BYOK Foundry gateway. The skill reads three environment variables — SCREENPIPETOKEN, ANTHROPICAPIKEY, FOUNDRYAPI_KEY — and uses each only to authenticate to the single endpoint its name implies. No other network destinations, no telemetry, no data egress to any third party.

Overview

Turn the user's own workflow history — captured passively by the local screenpipe daemon — into new skills. This skill is on-demand: the user invokes it with a time window, it queries screenpipe's local HTTP API, clusters repeated workflow patterns, compares each pattern against the existing skills in this repo, and produces a staged folder of proposals the user can review, edit, and promote.

When to Use This Skill

Invoke this skill when the user asks to:

"Analyze my last 4 hours / day / week and propose new skills."
"Look at what I've been doing and tell me what's not covered yet."
"Draft a skill from my recent workflow."
"Find composition recipes for workflows I repeat."

Do not invoke it for one-off questions about screenpipe itself, for real-time screen queries, or without an explicit user request — the skill analyzes sensitive local content and must stay explicitly user-triggered.

Privacy Posture

Screenpipe handles app/window filtering at capture time. Install a starter deny-list by copying references/screenpipe-config.yaml into the user's screenpipe config. Sensitive apps (password managers, messaging, banking) are never OCR'd in the first place.
Raw OCR never leaves the machine. scripts/fetch_window.py pulls data over localhost HTTP. scripts/cluster.py reduces the timeline to app/duration/title summaries. scripts/redact.py strips emails, API keys, bearer tokens, and phone numbers as defense-in-depth before any cluster summary reaches the LLM.
LLM backend defaults to local. The recommended setup is LM Studio running Gemma-4-31B-it — strong reasoning at a size that fits on most workstation GPUs, and no data ever leaves your machine. Cloud backends (claude, foundry) are opt-in and documented in config.yaml for users who explicitly want them. Detection and embeddings always run locally regardless of backend choice.
Dry-run mode (--plan) prints the exact timeline that will be analyzed before any LLM call.
TLS for localhost (optional, for corporate policy): see references/https-proxy.md for the Caddy pattern.

Prerequisites

1. Screenpipe daemon

Either install the official release or build from source. Either way the daemon binds HTTP on localhost:3030 by default.

From source (recommended if you want the CLI daemon without the desktop GUI):

git clone --depth 1 https://github.com/mediar-ai/screenpipe.git
cd screenpipe
cargo build -p screenpipe-engine --release
# System deps (macOS): cmake + full Xcode.app (not just Command Line Tools).
#   brew install cmake
#   # if xcodebuild plug-ins error: sudo xcodebuild -runFirstLaunch
./target/release/screenpipe doctor   # confirm permissions + ffmpeg
./target/release/screenpipe record --disable-audio --use-pii-removal

First run will prompt for macOS Screen Recording permission. Grant it and relaunch.

2. Screenpipe API token

The local API now requires bearer auth. Retrieve your token and export it:

export SCREENPIPE_TOKEN=$(screenpipe auth token)

(Or set screenpipe.token directly in config.yaml — env var is preferred since it keeps secrets out of version control.)

3. Python environment

Via pipenv from the repo root:

pipenv install httpx pyyaml sentence-transformers

The embedding model (sentence-transformers/all-MiniLM-L6-v2, ~80 MB) downloads on first run.

4. Local LLM (default path) — LM Studio

Install LM Studio.
Download Gemma-4-31B-it (or another strong reasoning model; adjust local.model in config.yaml).
Load it via the CLI for headless use (no GUI required):

lms load gemma-4-31b-it --context-length 131072 --gpu max -y
lms status   # confirm server running on :1234

5. Cloud LLM backends (optional, opt-in)

Only if you explicitly opt out of local:

claude: set ANTHROPICAPIKEY, flip backend: claude in config.yaml.
foundry: set FOUNDRYAPIKEY, flip backend: foundry, set foundry.endpoint to your corporate gateway URL.

Architecture

screenpipe daemon (user-installed)
        │  HTTP on localhost:3030
        ▼
scripts/fetch_window.py    → normalized timeline events
scripts/redact.py          → regex scrub (defense-in-depth)
scripts/cluster.py         → sessions + clusters (local only)
scripts/match_skills.py    → top-k vs existing 135 skills (local embeddings)
scripts/synthesize.py      → LLM judge: reuse / compose / novel
        │
        ▼
~/.autoskill/proposed/<timestamp>/        (default; override with --out)
  ├── report.md
  ├── composition-recipes/<name>/SKILL.md
  └── new-skills/<name>/SKILL.md

scripts/promote.py         → user-approved proposal → skills/<name>/

Workflow

The skill ships a unified CLI at scripts/autoskill.py with three subcommands:

python scripts/autoskill.py doctor   --config config.yaml --skills-dir ../
python scripts/autoskill.py run      --start ... --end ... --config config.yaml
python scripts/autoskill.py promote  --proposed ~/.autoskill/proposed/<ts> --skills-dir ../ --name <skill>

0. Preflight with doctor

Before a full run, verify every dependency in one shot:

python scripts/autoskill.py doctor \
  --config skills/autoskill/config.yaml \
  --skills-dir skills

The report covers config (backend choice valid), skills_dir (exists), screenpipe (reachable + authed), and llm (LM Studio serving or API key present). Non-zero exit on any failure, with the offending line marked error.

1. Run the pipeline

export SCREENPIPE_TOKEN=$(screenpipe auth token)
python scripts/autoskill.py run \
  --start "2026-04-17T00:00:00Z" \
  --end   "2026-04-17T23:59:59Z" \
  --config skills/autoskill/config.yaml \
  --skills-dir skills

Proposals land in ~/.autoskill/proposed// by default, keeping experimental output out of the skills repo. Pass --out PATH to override.

Internally:

Fetch — fetchwindow paginates screenpipe's /search endpoint, normalizes events to {ts, app, windowtitle, text, content_type}.
Redact — redact scrubs emails, API keys, bearer tokens, phones from OCR text and window titles as defense-in-depth over screenpipe's own PII removal.
Cluster — segmentsessions splits on idle gaps (default 10 min) and drops short sessions; clustersessions groups sessions by app-signature and keeps clusters of size minclustersize (default 2).
Match — loadskilldescriptions reads frontmatter from every SKILL.md in skills/; topkmatches ranks each cluster against all skills using local sentence-transformers embeddings (cosine similarity).
Synthesize — synthesize prompts the configured LLM backend to classify each cluster as reuse, compose, or novel and emit a SKILL.md body where appropriate.
Report — writes //report.md, plus new-skills//SKILL.md or composition-recipes//SKILL.md for each proposal.

Add --dry-run to stop after clustering; this skips the LLM (and the sentence-transformers load), writing only plan.md for inspection.

2. Review and promote

Open ~/.autoskill/proposed//report.md, edit drafts in place, delete anything you don't want. Then:

python scripts/autoskill.py promote \
  --proposed ~/.autoskill/proposed/2026-04-17T14-30-00 \
  --skills-dir skills \
  --name zotero-pubmed-helper

promote moves the directory into skills//, refusing to overwrite an existing skill. Exits non-zero with a friendly error if the proposal isn't found or the target already exists.

Configuration

See config.yaml for the full shape. Default values (local-first):

backend: local
local:
  endpoint: http://localhost:1234/v1   # LM Studio's Developer server
  model: Gemma-4-31B-it

screenpipe:
  url: http://localhost:3030           # or https://screenpipe.local via Caddy

cluster:
  min_session_minutes: 5
  idle_gap_minutes: 10
  min_cluster_size: 2

To opt into a cloud backend:

backend: claude                         # or foundry
claude:
  model: claude-opus-4-7

Composition recipes vs new skills

compose: the LLM judged that chaining existing skills covers the workflow. The emitted SKILL.md is intentionally thin — frontmatter + a "Workflow" section that invokes existing skills in order. The same agent runtime that discovered the skill can then invoke it end-to-end.
novel: no combination of existing skills covers it. A fuller SKILL.md is drafted, still following repo conventions (frontmatter, Overview, When to Use, Workflow). The user should always review new-skill drafts before promoting.

Testing

The skill is covered by a small pytest suite at tests/autoskill/ in the repository root. Each script is unit-tested in isolation with dependency injection (mock HTTP transport, stub backend, stub embedder):

python -m pytest tests/autoskill -v

More skills from K-Dense-AI/scientific-agent-skills

AadaptyvHow to use the Adaptyv Bio Foundry API and Python SDK for protein experiment design, submission, and results retrieval. Use this skill whenever the user mentions Adaptyv, Foundry API, protein binding assays, protein screening experiments, BLI/SPR assays, thermostability assays, or wants to submit protein sequences for experimental characterization. Also trigger when code imports `adaptyv`, `adaptyv_sdk`, or `FoundryClient`, or references `foundry-api-public.adaptyvbio.com`.
AaeonThis skill should be used for time series machine learning tasks including classification, regression, clustering, forecasting, anomaly detection, segmentation, and similarity search. Use when working with temporal data, sequential patterns, or time-indexed observations requiring specialized algorithms beyond standard ML approaches. Particularly suited for univariate and multivariate time series analysis with scikit-learn compatible APIs.
AalphagenomeLook up precomputed AlphaGenome Atlas effects for any GRCh38 single-nucleotide variant (AVI score with Phred and 18 SHAP feature attributions, plus raw and quantile scores for RNA-seq, DNase, ATAC, ChIP-TF, ChIP-histone, CAGE, PRO-cap, splicing, polyadenylation and contact-map tracks), score variants or scan windows on demand with the AlphaGenome model for human and mouse (variant scoring, in silico mutagenesis, REF-versus-ALT track prediction), and build Atlas website deep links. Use when the user mentions AlphaGenome, AlphaGenome Atlas, AVI or AlphaGenome Variant Impact, DeepMind variant effect prediction, or wants to prioritise or mechanistically interpret non-coding, regulatory, splicing, enhancer, promoter, or chromatin-accessibility effects of SNVs from a VCF, credible set, or region. Research use only; not a clinical tool.
Aanalytical-method-validationPlan, execute, and document validation, verification, and transfer of analytical procedures under the governing framework - ICH Q2(R2) and Q14, USP <1220>/<1225>/<1226>, ICH M10 bioanalytical, CLSI EP, or ISO/IEC 17025. Use for HPLC, LC-MS/MS, GC, CE, ICP-MS, dissolution, qNMR, qPCR, NIR, and ligand binding or cell-based assays whenever the question is whether a procedure is fit for its intended purpose. Triggers include "method validation", "analytical method validation", "AMV", "validation protocol", "acceptance criteria", "linearity", "reportable range", "accuracy and precision", "repeatability", "intermediate precision", "recovery", "LOD", "LOQ", "detection limit", "quantitation limit", "specificity", "robustness", "method transfer", "method comparison", "Deming", "Passing-Bablok", "Bland-Altman", "equivalence testing", "OOS investigation", "ICH Q2", "Q2(R2)", "Q14", "USP 1225", "ICH M10", "incurred sample reanalysis", "ISR", "CLSI EP", and any request to show that an assay works.
AanndataData structure for annotated matrices in single-cell analysis. Use when working with .h5ad files or integrating with the scverse ecosystem. This is the data format skill—for analysis workflows use scanpy; for probabilistic models use scvi-tools; for population-scale queries use cellxgene-census.
AarborAutonomously improve a real artifact (code, training recipe, agent harness, data pipeline, prompt) against an objective and an evaluator, using Hypothesis Tree Refinement (HTR) from the Arbor paper. Use this whenever someone wants to iteratively optimize something over many experiments without overfitting — e.g. "get my model's eval score up", "improve this agent/harness", "tune this pipeline", "beat the baseline on this benchmark", "run a search over approaches and keep the best", "do an MLE-bench / Kaggle-style optimization", or any long-horizon "make this artifact better and don't just memorize the dev set" task. Trigger it even when the user doesn't say "Arbor" or "hypothesis tree" but describes repeated experiment-and-evaluate loops, branching exploration of competing ideas, or worries about a dev/test gap. Runs Claude itself as the coordinator with subagent executors in isolated git worktrees; for the standalone `arbor` CLI tool see references/arbor-upstream.md.
AarboretoInfer gene regulatory networks (GRNs) from gene expression data using scalable algorithms (GRNBoost2, GENIE3). Use when analyzing transcriptomics data (bulk RNA-seq, single-cell RNA-seq) to identify transcription factor-target gene relationships and regulatory interactions. Supports distributed computation for large-scale datasets.
AastropyCore Python library for astronomy and astrophysics workflows that need Astropy APIs, including units/quantities, coordinates, FITS I/O, tables, time systems, WCS, and cosmology. Use when implementing or debugging astronomical data analysis code with Astropy.
Abenchling-integrationBenchling Python SDK and REST API integration for registry entities, inventory, ELN entries, workflows, Benchling Apps, and Data Warehouse queries. Use when automating lab data with benchling-sdk or the v2 API.
Abgpt-paper-searchSearch scientific papers and retrieve structured experimental data extracted from full-text studies via the BGPT MCP server. Returns 25+ fields per paper including methods, results, sample sizes, quality scores, and conclusions. Use for literature reviews, evidence synthesis, and finding experimental details not available in abstracts alone.
AbidsUse this skill when working with Brain Imaging Data Structure (BIDS) datasets: organizing neuroscience and biomedical data (MRI, EEG, MEG, iEEG, PET, microscopy, NIRS, motion capture, EMG, MR spectroscopy, behavioral), querying BIDS layouts, validating compliance, converting DICOM to BIDS, writing metadata sidecars, or creating BIDS derivatives.
AbiopythonComprehensive molecular biology toolkit. Use for sequence manipulation, file parsing (FASTA/GenBank/PDB), phylogenetics, and programmatic NCBI/PubMed access (Bio.Entrez). Best for batch processing, custom bioinformatics pipelines, BLAST automation. For quick lookups use gget; for multi-service integration use bioservices.

All agent skills → · MCP servers