Skip to content
Open
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension


Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
34 changes: 34 additions & 0 deletions citation-context-reconciler/README.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,34 @@
# Citation Context Reconciler

`citation-context-reconciler` is a focused SCIBASE AI research assistant module for issue #16. It helps authors and reviewers catch citation drift before a manuscript leaves the workspace: sources cited as evidence are checked against the actual source context, effect direction, method match, population match, recency, and reproducibility artifacts.

The slice is intentionally narrower than a broad copilot. It answers a practical question during pre-submission review: "Are these citations being used honestly enough for a reviewer, journal, funder, or replication team to trust the claim?"

## What It Does

- Reviews manuscript claims against cited source metadata.
- Detects direct blockers such as contradictory cited effects, citation-intent mismatch, retracted citations, missing sources, and effect-direction mismatch.
- Raises warnings for mixed evidence, stale citations, population/method context gaps, and missing raw data/code/protocol evidence.
- Generates reviewer comments and author revision tasks.
- Scores reproducibility support from raw data, code, and registered protocol availability.
- Ranks research opportunities from contradictions, lab capabilities, and researcher interests.
- Exports a deterministic assistant packet with an audit digest for handoff and version history.

## Files

- `src/citation-context-reconciler.js` - dependency-free reconciliation engine.
- `sample/citation-context-packet.json` - synthetic manuscript/source packet.
- `test/citation-context-reconciler.test.js` - executable behavioral tests.
- `scripts/demo.js` - writes `docs/citation-context-report.json` and `docs/demo.svg`.
- `scripts/render-demo-video.m` - macOS-native MP4 renderer for `docs/demo.mp4`.
- `docs/requirement-map.md` - issue #16 capability mapping.

## Validation

```sh
npm test
npm run demo
npm run demo:video
```

The demo uses only local synthetic data and does not call an AI provider or external service.
Loading