refresh-research

To keep research documentation from going stale, you can re-fetch eligible upstream sources and apply detected changes to synthesis files using a staleness audit and refresh process.

The best way to audit a knowledge base for outdated sources is running an audit-only dry run, which reports staleness without modifying files, or by triggering a refresh based on age thresholds.

Versioned source storage works by saving every re-fetched upstream update as a new dated raw file, ensuring historical data is never overwritten during synthesis updates.

Yes, you can refresh specific categories in your research corpus, as the process supports single files, whole categories, full corpora, and staleness-threshold refreshes.

Yes, you need Python 3.10 or higher to run a staleness audit on GitHub documentation and re-fetch eligible repo and docs sources for your knowledge base.

Refreshing upstream sources will not overwrite your existing analysis; the process applies surgical synthesis updates that preserve your original notes while logging changes in an Update Log.