R0023/2026-03-25/Q001/SRC01
The Prompt Report — most comprehensive systematic survey of prompt engineering techniques
Source
| Field |
Value |
| Title |
The Prompt Report: A Systematic Survey of Prompting Techniques |
| Publisher |
arXiv (preprint) |
| Author(s) |
Sander Schulhoff et al. (31 authors from OpenAI, Microsoft, Google, Princeton, Stanford, UMD) |
| Date |
2024-06-06 (v1), 2025-02-26 (v6) |
| URL |
https://arxiv.org/abs/2406.06608 |
| Type |
Systematic review / meta-analysis |
Summary
| Dimension |
Rating |
| Reliability |
High |
| Relevance |
Medium |
| Bias: Missing data |
Low risk |
| Bias: Measurement |
N/A |
| Bias: Selective reporting |
Low risk |
| Bias: Randomization |
N/A — not an RCT |
| Bias: Protocol deviation |
N/A — not an RCT |
| Bias: COI/Funding |
Low risk |
Rationale
| Dimension |
Rationale |
| Reliability |
PRISMA-based systematic review of 1,565 papers. Multi-institutional authorship with major AI labs. Six revisions over 8 months demonstrate ongoing quality control. |
| Relevance |
Provides the taxonomic framework for prompt engineering techniques but is primarily a survey rather than a study of counterproductive effects specifically. Medium relevance because it catalogs what exists rather than evaluating what fails. |
| Bias flags |
Low risk across the board. Multi-institutional authorship reduces single-entity bias. PRISMA methodology enforces transparent search and selection processes. |
| Evidence ID |
Summary |
| SRC01-E01 |
Identification of 58 text-based prompting techniques categorized into 6 groups, establishing the taxonomy against which counterproductive findings can be mapped |