literature-retriever
Skill10 starsChanged 6 days agoMITView on GitHub
Searches arXiv for a keyword phrase and stores the result as JSON in `process/`. Returns metadata only — titles, authors, identifiers, abstracts — and never downloads or redistributes the papers themselves.
In this repository22 other files it gives its agents.
- AGENTS.mdAGENTS.md
- .github/copilot-instructions.mdCopilot instructions
- .cursor/rules/chelate-shell.mdcCursor rule
- 3d-modelerSkill
- ai-trace-removerSkill
- bibtex-normalizerSkill
- capstone-labmateSkill
- conflict-detectorSkill
- data-profilerSkill
- document-evolution-trackerSkill
- formula-validatorSkill
- fulltext-extractorSkill
- latex-exporterSkill
- longitudinal-analysisSkill
- ml-auditorSkill
- multi-dim-managerSkill
- quantitative-validatorSkill
- reference-auditorSkill
- sci-visualizerSkill
- security-guardSkill
- task-organizerSkill
- video-trackerSkill
The file
We split this skill into the rules it gives an agent
3 rules. About 0 are common, 3 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈3
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 25
Run this exact form — do not read the source tree:
- Distinctive: line 53
Metadata only: No PDF is downloaded.
- Distinctive: line 54
The API's answer is reported as-is: No invented hits.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
literature-retriever
Queries the arXiv API and normalises the response into one JSON file per search. It does not rank by relevance for you beyond the API's own sort.
- Stage:
ingest· Sandbox: no · Dependencies:arxiv
Quick Execution
Run this exact form — do not read the source tree:
python run.py --root <WS> --action search --query "topological insulator"--max-resultsand--sort-by {relevance,lastUpdatedDate,submittedDate}default toassets/config.yaml.- Results land in
<WS>/process/literature/, the gate in<WS>/process/gates/.
When to use
- A topic needs a first pass of candidate papers.
- A keyword search result must be captured as structured JSON for later stages.
When NOT to use
- You need the full text — use
fulltext-extractoron a PDF you have. - The citation must be verified — that is
reference-auditor's job.
Actions
--action | Does | Extra arguments |
search | arXiv keyword search → JSON | --query "..." [--max-results N] [--sort-by ...] [--outdir <dir>] |
Hard constraints
- Metadata only. No PDF is downloaded.
- The API's answer is reported as-is. No invented hits.
Outputs
<WS>/process/literature/<slug>_<ts>.json— the normalised result set.<WS>/process/gates/literature-retriever_<ts>.json— the gate result.
Failure handling
- A network failure fails the gate with the HTTP error.
arxivmissing →UNVERIFIEDwith an install hint.
Limitations
- arXiv only; other indexes are out of scope.
- Ranking is the API's, not a relevance model.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.