skill-rss-aggregator · RSS Feed Aggregator
RSS/Newsletter aggregator: subscribes to a batch of bloggers', media, and newsletters' RSS/Atom feeds, pulls the latest entries, filters by keywords and time window, dedupes, sorts by time, and produces topic and news digests.
Pure-standard-library parsing script (zero third-party dependencies), accepting single URLs or a feed list file. Digests list titles, sources, dates, excerpts, and links — never full text. Failed feeds are skipped with a notice, never aborting the whole aggregation.
Example invocation: "Summarize the last 7 days of updates from these blogs I follow."
Full brief
Positioning
skill-rss-aggregator aggregates updates from user-chosen subscription feeds only. For web-wide trending topics use skill-trending-topics; for industry news use skill-news-intelligence.
Core capabilities
- Standard-library parsing:
scripts/rss_digest.pyparses RSS 2.0 / Atom with stdlib (XML parsing + HTML cleaning + date-format tolerance), zero third-party dependencies. - Multi-source aggregation: single URL or feed list file (one URL per line,
#comments). - Keyword & time-window filters: keep entries whose title/excerpt matches keywords; keep only the last N days; entries without pubDate are never wrongly dropped by the time window (kept, sorted last).
- Dedup & sort: cross-source dedup, time-sorted.
- Markdown digest output: title/source/date/excerpt/link, with
-oto save. - Copyright-respecting: digests never fetch full text.
- Fault-tolerant: failed feeds skipped with a notice, never aborting aggregation.
- No-RSS guidance: sites without RSS can get feeds via RSSHub; prompts the user to set it up.
Workflow
- Get subscription feeds (single URL or list file)
- Script fetch + parse (
fetch --feeds feeds.txt --keyword ... --since N -o ...) - Keyword filter + dedup + time sort
- Output the Markdown digest
Inputs & outputs
| Input | Notes |
|---|---|
| Feeds | Required — --url single or --feeds list file |
| Keywords | Optional, comma-separated; keep matching entries |
| Time window | Optional — last N days only |
Output: digest Markdown (title/source/date/excerpt/link) or JSON.
Boundaries with adjacent skills
| Skill | Lane |
|---|---|
| skill-rss-aggregator (this) | Updates from user-chosen subscription feeds |
| skill-trending-topics | Web-wide trending topics |
| skill-news-intelligence | Industry news |
Fit
- Daily/weekly news briefings (scheduled runs)
- Tracking followed bloggers/newsletters for writable topics
- Digest → skill-topic-evaluator for feasibility → skill-content-matrix for scheduling
Before you start
- Prepare the feeds' RSS/Atom URLs (RSSHub can generate feeds for sites without RSS).
- With a profile, default keywords come from the account's niche; only on-position updates are pushed.
skill-rss-aggregator is part of the Aiglade Skill library. Invoke it from the Aiglade chat box in plain language.