you-read-i-read

Add a paper or blog from a URL. Fetches, deep-summarizes, files to to-read.

Mark a paper archived (decision: not going to read).

Mark a paper read (when finished outside a /read session).

Show papers per the user's filters: $ARGUMENTS

Start an interactive reading session. With no args, agent recommends. With <id>, that paper. With keywords, search.

Commit and (with confirmation) push the private data repo.

List the to-read queue, ranked. Optional query filters.

Track a research group's publications.

Track a topic across arXiv, Semantic Scholar, HN, and (opt-in) Twitter/LinkedIn.

Sweep tracked sources for new papers since last check; triage with the user.

Specialized agent for fetching a single paper / blog / PDF URL and returning a structured 7-section markdown summary (motivation, problem statement, challenge, insight, tech details, evaluation, takeaways). Use for parallel summarization fanout from paper-curate during /add and from /update [r]ead-now paths. Persists raw fetched content under papers/raw/ when summary.store_raw_content is true.

Use when adding a paper, blog, or report by URL, tracking a research group or topic, or sweeping tracked sources for new candidates. Triggers include "/add <url>", "/track-group", "/track-topic", "/update", "add this paper", "follow this group", "watch this topic", "any new papers since last week", "summarize and save this".

Use when listing the to-read queue or any paper subset, editing paper metadata (status, labels, one-liner), changing read/archive status outside a /read session, viewing tracked groups/topics, syncing the data repo to git, or running first-run setup. Triggers include "/todo", "/list", "/done", "/archive", "/sync", "show my queue", "what am I tracking?", "remove this paper", "I finished reading X", "mark X read", "archive that paper", "push my data", "skip this one not going to read it".

Use when starting an interactive reading session over a paper. Triggers include "/read", "/read <id>", "/read <keywords>", "let's read something", "I want to read about <topic>", "let's read X", "show me a paper to read". The agent picks from the to-read queue, loads the structured summary, walks through it menu-driven, drills into raw text on request.