| name | assess-citations |
| description | Assesses citations in the paper draft against source papers in literature/. Runs as a background agent. |
| context | fork |
| agent | citation-assessor |
| disable-model-invocation | true |
When invoked, ALWAYS use the Agent tool to run the agent: '/citation-assessor'
Do NOT execute the agent's task inline.
The agent walks through every citation and placeholder in pipeline/paper_draft.md, checks each one against the source papers in literature/, and produces a structured report with proposed actions (KEEP / REVISE / REPLACE / REMOVE). The result is written to pipeline/citation_report.md.
Before proceeding: is there anything specific I should know — citations you're already aware are wrong, new papers you've added to literature/, or sections to prioritize? If not, just say go.