Research Cite Command

Generate properly formatted, policy-compliant citations from the research corpus.

Instructions

When invoked, generate correct citations:

1. Locate Source

   • Search .aiwg/research/sources/ and .aiwg/research/findings/ for specified reference
   • Match by REF-XXX identifier, author name, or keyword
   • If multiple matches, present options for user selection

2. Load Metadata

   • Extract frontmatter: ref_id, title, authors, year, DOI, source_type
   • Load quality assessment if available
   • Determine GRADE level

3. Generate Citation

   • Format @-mention citation with full path
   • Include page numbers if specified
   • Apply GRADE-appropriate hedging language
   • Include quality level annotation

4. Output Formats

   • inline - Ready to paste into markdown with GRADE-appropriate language
   • bibtex - BibTeX format for reference management
   • reference - Full reference section entry
   • apa - APA 7th edition format
   • chicago - Chicago Manual of Style format

Arguments

• [ref-id or keyword] - REF-XXX identifier or search keyword (required)
• --format [inline|bibtex|reference|apa|chicago] - Output format (default: inline)
• --page [n] - Include page reference
• --quote "[text]" - Include direct quote with citation
• --hedging [suggest|none] - Include hedging language suggestions (default: suggest)

Examples

# Basic inline citation
/research-cite REF-022

# Citation with page reference
/research-cite REF-022 --page 4

# Citation with quote
/research-cite REF-022 --page 4 --quote "deliberate decision making"

# BibTeX format for bibliography
/research-cite REF-022 --format bibtex

# Search by keyword
/research-cite "autogen multi-agent" --format inline

# Multiple citations for bibliography
/research-cite REF-022 REF-001 REF-013 --format reference

Expected Output

Inline Format (Default)

/research-cite REF-022 --page 4

Output:
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

Citation (inline format):

  According to @.aiwg/research/findings/REF-022-autogen.md (p. 4),
  AutoGen enables "flexible conversation patterns" through multi-agent
  collaboration.

GRADE Assessment: MODERATE
  Source: arXiv preprint (peer-reviewed conference version available)
  Quality: Moderate (single study, not systematic review)

Recommended Hedging Language:
  ✓ APPROPRIATE: "Research suggests...", "Evidence indicates..."
  ✗ TOO STRONG: "Research proves...", "Evidence demonstrates..."
  ✗ TOO WEAK: "Some sources claim...", "Anecdotal reports suggest..."

Usage Example:
  Multi-agent conversation patterns suggest flexible agent orchestration
  is feasible for complex workflows (@.aiwg/research/findings/REF-022-autogen.md,
  p. 4), though scalability limits require further investigation (GRADE: MODERATE).

BibTeX Format

/research-cite REF-022 --format bibtex

Output:
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

@article{wu2023autogen,
  title={AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversation},
  author={Wu, Qingyun and Bansal, Gagan and Zhang, Jieyu and Wu, Yiran and Zhang, Shaokun and Zhu, Erkang and Li, Beibin and Jiang, Li and Zhang, Xiaoyun and Wang, Chi},
  journal={arXiv preprint arXiv:2308.08155},
  year={2023},
  doi={10.48550/arXiv.2308.08155},
  note={GRADE: MODERATE - arXiv preprint}
}

Reference Format

/research-cite REF-022 --format reference

Output:
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

Wu, Q., Bansal, G., Zhang, J., Wu, Y., Zhang, S., Zhu, E., Li, B.,
Jiang, L., Zhang, X., & Wang, C. (2023). AutoGen: Enabling Next-Gen
LLM Applications via Multi-Agent Conversation. *arXiv preprint*
arXiv:2308.08155. https://doi.org/10.48550/arXiv.2308.08155

GRADE: MODERATE
Internal Reference: @.aiwg/research/findings/REF-022-autogen.md
PDF: @.aiwg/research/sources/REF-022.pdf

GRADE-Based Hedging

Citations automatically include hedging guidance based on evidence quality:

GRADE Level	Appropriate Language	Avoid
HIGH	"demonstrates", "shows", "confirms"	Uncertainty language
MODERATE	"suggests", "indicates", "supports"	Definitive claims
LOW	"limited evidence", "preliminary findings"	Strong claims
VERY LOW	"anecdotal", "exploratory", "reports suggest"	Confident claims

Batch Citation

Generate bibliography for multiple sources:

/research-cite REF-001 REF-013 REF-022 REF-057 --format reference

Output:
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

## References

Feng, Y., et al. (2024). The Landscape of Emerging AI Agent Architectures...
  GRADE: MODERATE - @.aiwg/research/findings/REF-001-production-agentic.md

Hong, S., et al. (2023). MetaGPT: Meta Programming for Multi-Agent Systems...
  GRADE: HIGH - @.aiwg/research/findings/REF-013-metagpt.md

Wu, Q., et al. (2023). AutoGen: Enabling Next-Gen LLM Applications...
  GRADE: MODERATE - @.aiwg/research/findings/REF-022-autogen.md

Schmidgall, S., et al. (2024). Agent Laboratory: Using LLM Agents...
  GRADE: HIGH - @.aiwg/research/findings/REF-057-agent-laboratory.md

Citation Validation

All citations are validated against:

• Source exists in .aiwg/research/
• Metadata is complete (title, authors, year)
• DOI/identifier is valid
• GRADE level is assessed
• Hedging language matches quality level
• Page references are within paper bounds (if PDF available)
• Quotes are exact (if --quote provided)

Integration with Writing

When writing documentation:

1. Use /research-cite REF-XXX to get citation
2. Copy inline format into document
3. Use suggested hedging language
4. Add @-mention for traceability
5. Include GRADE level in internal notes

References

• @$AIWG_ROOT/agentic/code/frameworks/research-complete/agents/citation-agent.md - Citation Agent
• @$AIWG_ROOT/src/research/services/citation-service.ts - Citation formatting
• @$AIWG_ROOT/agentic/code/frameworks/sdlc-complete/rules/citation-policy.md - Citation policy requirements
• @$AIWG_ROOT/agentic/code/frameworks/sdlc-complete/schemas/research/frontmatter-schema.yaml - Source metadata
• @.aiwg/research/docs/grade-assessment-guide.md - GRADE levels