OpenSkillIndex← back to search
claude-codedata

Claude-skill-registry consume-feed

install
source · Clone the upstream repo
git clone https://github.com/majiayu000/claude-skill-registry
Claude Code · Install into ~/.claude/skills/
T=$(mktemp -d) && git clone --depth=1 https://github.com/majiayu000/claude-skill-registry "$T" && mkdir -p ~/.claude/skills && cp -r "$T/skills/data/consume-feed" ~/.claude/skills/majiayu000-claude-skill-registry-consume-feed && rm -rf "$T"
manifest: skills/data/consume-feed/SKILL.md
tags
#rss-ingestion#data-pipeline#knowledge-graph#feed-processing
source content

Consume Feed Skill

A robust ingestion engine for upstream data sources.

Phase 1: RSS (Implemented)

  • "Pull the feeds now" -> 
    ./run.sh run --mode manual
  • "Add this RSS feed <url>" -> 
    ./run.sh sources add rss --url <url>
  • "Check feed ingest health" -> 
    ./run.sh doctor

Phase 2: GitHub & NVD (Aspirational)

  • "Add GitHub repo <owner>/<repo>" -> 
    sources add github --repo <owner>/<repo>
  • "Track NVD for <keyword>" -> 
    sources add nvd --query <keyword>

Usage

Run Ingestion

# Run nightly crawl (all sources, respect intervals)
./run.sh run --mode nightly

# Run specific source immediately
./run.sh run --source <key>

Manage Sources

# List all
./run.sh sources list

# Add RSS
./run.sh sources add rss --url "https://github.blog/feed/"

Diagnosis & Initialization

# Health check
./run.sh doctor

# Force initialize search views and indexes
./run.sh doctor --init

Resilience

  • Uses exponential backoff and jitter for all network requests.
  • Persists checkpoints (ETags, Timestamps) to resume efficiently.
  • Reuses Memory skill connection for stable, shared database access.