Detect repeated research workflows from local screenpipe activity and draft skill proposals.
npx skills add https://github.com/dralkh/seerai --skill autoskill-dralkh
Please help me install this Agent Skill. Skill: autoskill Source: https://github.com/dralkh/seerai/tree/main/skills/autoskill Command: npx skills add https://github.com/dralkh/seerai --skill autoskill-dralkh
Autoskill turns your repeated research work into new Skill drafts, so you can spot workflow patterns you already perform and convert them into reusable automations instead of rebuilding them manually.
Ask autoskill to analyze a recent time window from your local screenpipe history and draft skill proposals for any repeated workflows it finds.
High-intent search queries and answers about installing and using this skill.
Research workflows can be drafted into reusable skills by analyzing local screenpipe activity to detect repeated patterns and synthesizing them into skill proposals. Autoskill groups recurring desktop sessions, filters out one-off tasks, and generates reviewable drafts.
Workflow detection reads local screenpipe history and groups repeated app and window-title patterns to identify recurring research tasks. It filters out short or one-off sessions, clustering redacted timelines to isolate genuine repeated workflows like literature reviews or citation management.
Local screenpipe access is required to detect and analyze research workflows. Autoskill reads your local screenpipe history to identify repeated patterns, keeping detection, clustering, and embeddings processing entirely local to preserve privacy.
Privacy is preserved by redacting sensitive text before LLM synthesis during skill generation. Autoskill keeps detection, timeline clustering, and semantic embeddings local, ensuring only redacted data is processed by the LLM.
Detected workflows are matched against existing skill descriptions using semantic comparison. Autoskill evaluates whether to reuse an existing skill, compose multiple skills together, or draft a new candidate skill based on the semantic similarity results.
Zotero-centered literature review tasks can be automated by capturing your repeated PubMed searches, Zotero note reading, and draft writing via screenpipe. Autoskill analyzes these recurring desktop activities to generate a reviewable skill proposal for that specific workflow.
handsomelong922Generate embeddings for semantic similarity, clustering, and retrieval with PyTorch.
dkmakerRetrieve coding documentation and persist structured findings to research files.
jasrajtulsiManage Zotero libraries via the Zotero Web API for CRUD, search, and export.
galeepAnalyze screen activity patterns to suggest new scientific skills or compositions.
