AI Agent Workflow Collection

Discover important work the radar does NOT yet track, by iterating the discovery-venue list in SOURCES.md every run — significance-first: read the top / most-attention items in each venue REGARDLESS of topic, not by searching tracked axes. Surfaces off-axis significance and candidate new axes. Use for the daily exploration slot, after the sweeps.

Sweep the official AI lab and big-tech AI blogs on EVERY daily run (not rotated) so no open-weight release, infrastructure or safety announcement is missed. Prefer RSS/Atom feeds over HTML pages to survive JavaScript-rendered indexes. Use at the start of every daily scan, before the rotating exploration slot.

Safely update TRENDS.md (the AI Radar ledger): append evidence, move stages, manage observationqueue, sourcerotation and strategynotes without breaking the file contract. Use every time TRENDS.md is edited, before committing radar updates, and when promoting or archiving trends.

Sample social/community sources (Reddit, Hacker News, YouTube, Hugging Face, and best-effort X/Instagram) for sentiment and leading-indicator buzz on every daily run. Social is an INTAKE LANE ONLY — it can feed observationqueue (unverified) and a community-pulse note, but NEVER becomes trend evidence. Use after the lab sweep and arXiv scan.

Regenerate README.md, the repo's landing page and only output surface: badges, a 3-4 bullet "since last scan" digest with inline links, a clickable trend table, the study shelf, and an output map. Use after every change to TRENDS.md, in the same commit. README.md is fully derived and never edited by hand: if it disagrees with TRENDS.md, TRENDS.md wins — regenerate.

Watch GitHub every daily run for behind-the-scenes movement — across watched REPOS (releases, merged PRs, hot issues), watched PROFILES/USERS (what a key author ships next), and FORK TREES (notable forks up to depth 3, scored by a notability metric). Releases and merges are citable artifacts; profile/issue/fork movement is a queue signal until it lands. Default to GitHub's public Atom feeds + Tavily; the REST API is optional. Use during the daily scan, after the lab sweep. Source lists live in SOURCES.md → "GitHub watch".

Compute the radar's self-calibration: weekly funnel metrics (queue dynamics, exploration compliance, off-axis rate), a monthly hit/miss retrospective against what actually became big, and up to 3 curator proposals. Use during every weekly recalibration, after the source-strategy review; results go into the append-only logs/calibration.md and into the weekly report.

Repair a source or tool that keeps failing — a broken/moved/stale feed, a 404, an empty result, a parse/JSON error — instead of logging "degraded" forever. Diagnose the failure class, find a working access path, and record the repaired method in SOURCES.md so future runs don't repeat it. Generalizes to any future source/tool breakage. Use whenever the lab sweep, pulse, or repo-watch hits a source that failed the same way before.

Verify a signal or evidence candidate for the AI Radar ledger: open primary sources, pin exact dates and authors, check source independence, and classify the result as citable evidence or an unverified observationqueue item. Use whenever adding evidence to TRENDS.md, promoting a queue item, or fact-checking a claim found in search results.

Extract clean markdown or text from specific URLs via the Tavily CLI. Use this skill whenever a URL must be opened to be cited — papers, changelogs, release pages, blog posts, docs — including JavaScript-rendered pages. Processes up to 20 URLs per call and supports query-focused chunking for long pages.