Use when the user wants to find related work, survey a research area, identify literature gaps, or discover open-source implementations. Triggers on phrases like "find papers on", "related work", "literature review", "what papers exist", "open source implementation", or "papers with code".
SkillScore77Maintained
Automatic
◆Research pack
Last verified 8/21/2026
What it changes
Passed 17 of 22 test cases with the skill, and 10 of 22 without it.
Without skill45%
With skill77%
Checks it passed that the model alone missed
The coverage matrix assigns problem aspects to rows.
The topic summary overview is limited to 2 to 3 sentences.
The coverage matrix assigns existing approaches to columns.
and 8 more in the full report
8 fixed9 passed either way4 still failing1 broke
Pass-rate lift
+32 pp
Tokens
+10%
Agent turns
0%
Response time
-23%
Measured 8/21/2026v1 · latest is v217 of 22 cases passed with skill
Tokens, turns and time describe resource use, not a dollar price.
This benchmark measured an earlier version of the skill.
A newer attempt on gemini-3.8-flash on 10/20/2026 did not produce a publishable comparison (not comparable). The snapshot above is the previously measured run.
See it in action
One of 8 test cases the skill fixed. Same model, same prompt, with and without the skill.
1 of 8
Prompt
Which four target database platforms should be searched when conducting direct key-term queries for a systematic review on agentic software engineering tools?
What was checked, without and with the skill
Checked
Without skill
With skill
The direct search target platforms include Google Scholar, Semantic Scholar, arXiv, and DBLP.
Missed
Passed
case-21 · gemini-3.6-flash · 8/21/2026 · answers are capped and redacted
What's inside
The instructions your model reads when the skill loads.
SKILL.md92 linessingle file
Literature Research Methodology
You are helping a researcher conduct systematic literature research. Follow this methodology to ensure thorough, accurate coverage.
Step 1: Scope Definition
Before searching:
Clarify the exact research question or topic boundary
Identify key terms and their synonyms (e.g., "content moderation" = "safety filtering" = "NSFW detection")
Inspect the JSON errors field and retain each paper's source ID and URL. Report unavailable providers rather than treating errors as an empty literature search. Vary search terms to catch different framings of the same concept. Use web search for proceedings and sources the CLI cannot access.
2b. Citation Chaining
From seed papers or initial results:
Forward chaining: run paper-search tool get_citing_papers <doi>
Backward chaining: run paper-search tool get_referenced_papers <doi>
Both accept a DOI or OpenAlex work ID, not a title. Inspect pagination and truncation metadata; neither command promises an exhaustive citation list. Read the paper's reference section to verify cited works.
This catches papers that use different terminology but address the same problem
2c. Venue Mining
Identify top venues for the topic (conferences, journals, workshops)
Check recent proceedings of these venues for relevant papers
Workshop papers often contain early-stage work not yet in main conferences
2d. Open Source Discovery
Search GitHub for implementations related to the topic