mcpField
S
sustainability-summary

sustainability-summary

github.com/tiangong-aiProductivity1157 installsUpdated 1w ago
Repository

About sustainability-summary

sustainability-summary is a skill for producing evidence-grounded sustainability digests from relevant RSS records, with optional full-text data. It is intended for agents handling date-based or custom-period reporting and cites claims with DOI and URL.

Key features

  1. 01Daily, weekly, monthly, and custom reports
  2. 02Retrieves only relevant RSS records
  3. 03Optional full-text context
  4. 04DOI and URL evidence citations

Use cases

  1. 01Create a weekly sustainability digest
  2. 02Summarize records for a custom date range
  3. 03Ground claims in RSS and full-text evidence

Install

Copy the command and run it in a terminal.

$npx skills add https://github.com/tiangong-ai/agent-skills --skill sustainability-summary

After installing, type/sustainability-summaryin Claude Code to confirm it loaded.

Details

From the project README

Sustainability Summary

Core Goal

  • Read only relevant (is_relevant=1) records from RSS DB.
  • Build compact RAG context from DOI-keyed entries.
  • Include optional enriched content from separate fulltext DB entry_content when available.
  • Let the agent synthesize final summary text with evidence anchors.

Triggering Conditions

  • Receive requests for daily/weekly/monthly sustainability digests.
  • Receive requests for custom date-range summaries.
  • Need evidence-grounded output from labeled RSS entries and enriched content.

Input Requirements

  • Required tables: feeds, entries (from sustainability-rss-fetch).
  • entries must be DOI-keyed and relevance-labeled.
  • Optional fulltext DB table: entry_content (from sustainability-fulltext-fetch).
  • RSS DB and fulltext DB must be different files.

Workflow

  1. Build retrieval context by time window.
export SUSTAIN_RSS_DB_PATH="/absolute/path/to/workspace-rss-bot/sustainability_rss.db"
export SUSTAIN_FULLTEXT_DB_PATH="/absolute/path/to/workspace-rss-bot/sustainability_fulltext.db"

python3 scripts/time_report.py \
  --rss-db "$SUSTAIN_RSS_DB_PATH" \
  --content-db "$SUSTAIN_FULLTEXT_DB_PATH" \
  --period weekly \
  --date 2026-02-10 \
  --max-records 120 \
  --max-per-feed 20 \
  --summary-chars 8192 \
  --fulltext-chars 8192 \
  --pretty \
  --output /tmp/sustainability-weekly-context.json

  1. Generate final summary from returned records + aggregates.

  2. Cite evidence using DOI + URL for key claims.

Time Window Modes

  • --period daily --date YYYY-MM-DD
  • --period weekly --date YYYY-MM-DD
  • --period monthly --date YYYY-MM-DD
  • --period custom --start ... --end ...

Default Fields

  • doi,timestamp_utc,timestamp_source,feed_title,feed_url,title,url,summary,fulltext_status,fulltext_length,fulltext_excerpt

Configurable Parameters

  • --rss-db
  • --content-db
  • SUSTAIN_RSS_DB_PATH
  • SUSTAIN_FULLTEXT_DB_PATH
  • --period
  • --date
  • --start
  • --end
  • --max-records
  • --max-per-feed
  • --summary-chars
  • --fulltext-chars
  • --top-feeds
  • --top-keywords
  • --fields
  • --output
  • --pretty
  • --fail-on-empty

Error Handling

  • Missing required DOI-based tables: fail fast with setup guidance.
  • RSS DB and fulltext DB path collision: fail fast and require separate files.
  • Invalid date/time/field list: return parse errors.
  • Missing entry_content: continue in metadata-only mode.
  • Empty relevant set: return empty context; optional failure with --fail-on-empty.

References

  • references/time-window-rules.md
  • references/report-format.md

Assets

  • assets/config.example.json

Scripts

  • scripts/time_report.py

Comments

Markdown supported