citation-fixer

Audit and fix inline source citations across brain pages, resolving broken tweet references via X API.

Updated Aug 17, 2026
One-click install
npx skills add https://github.com/AMC-JTC/gbrain-1 --skill citation-fixer-amc-jtc
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: citation-fixer
Source: https://github.com/AMC-JTC/gbrain-1/tree/main/plugin/skills/citation-fixer
Command: npx skills add https://github.com/AMC-JTC/gbrain-1 --skill citation-fixer-amc-jtc

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Brain pages accumulate facts without proper inline citations, and tweet references often lack actual URLs, making claims unverifiable. This Skill scans every page, flags missing or malformed citations, rewrites them to the standard format, and resolves broken tweet references into deterministic x.com links. ## Core Features & Use Cases - Citation auditing: Scans all brain pages for missing, malformed, or incomplete [Source: ...] citations and reports counts of scanned, fixed, and remaining issues. - Tweet reference resolution: Detects tweet mentions without URLs, searches the host's X API integration by handle and quoted text, verifies matches, and patches pages with deterministic https://x.com/<handle>/status/<id> links including engagement metrics. - Batch mode with rate limiting: Sweeps up to ~50 pages per run, prioritizes recently updated pages, batch-commits every 10-20 pages, and tracks progress in a JSON state file. - Use Case: Run a weekly cron sweep so that every page ingested from meetings, emails, or tweets ends with verifiable citations instead of hand-wavy attributions. ## Quick Start Ask the agent to fix citations on a specific brain page or run a citation audit across recently updated pages.

Frequently Asked Questions about citation-fixer

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I fix broken citations across multiple pages?

Run the citation-fixer in batch mode, which scans pages for missing or malformed [Source: ...] citations and rewrites them to the standard format. It processes about 50 pages per run and commits every 10-20 pages to avoid losing progress.

How are tweet references without URLs resolved?

The skill extracts the handle, quoted text, and approximate date from each broken reference, then searches the host's X API integration. Verified matches are patched with deterministic https://x.com/<handle>/status/<id> links plus engagement metrics.

Does citation-fixer delete facts that have no source?

No. Facts without citations are flagged in the audit report as remaining gaps, never removed or given invented citations. The skill explicitly prohibits fabricating sources or deleting uncitable content.

Can other skills call citation-fixer automatically?

Yes. Skills like enrich or media-ingest can invoke citation-fixer before committing pages to validate output. It can also run manually on a single page or as a recurring weekly cron job.

What rate limits apply when resolving tweets via the X API?

The skill respects the host's X API tier limits, targeting roughly 1-3 API calls per page for search and verification. Batch runs are capped near 50 pages to stay within quota.