| name | arxiv-search |
| description | Academic paper search via arXiv. Returns ranked paper results per query envelope. |
| version | 0.1.0 |
| purity | pure-data |
| runtime | research-runner |
| command | npx @maintainabilityai/research-runner@~0.1.64 skill-arxiv-search |
arXiv Search Skill
Call when the agent needs peer-reviewed / preprint academic sources for research evidence. PURE-data: no LLM inside; the parent agent crafts queries (formal CS / domain terms — see prompt-packs/looking-glass/research/query-plan.md anchors).
Inputs (stdin JSON)
| Field | Type | Required | Default | Description |
|---|
queries | string[] | yes | — | Formal-domain queries: 3-6 word technical phrases, no boolean operators, no recency year (e.g. "named entity disambiguation knowledge graph"). |
maxResults | integer | no | 10 | Per-query result cap. |
Outputs (stdout JSON)
| Field | Type | Description |
|---|
envelopes | {query, status, count, error?}[] | Per-query envelope. |
results | ProviderResult[] | Flattened ranked list with abstract field per paper. |
Invocation
echo '{"queries":["named entity disambiguation knowledge graph"]}' \
| npx @maintainabilityai/research-runner@~0.1.64 skill-arxiv-search
Implementation
Thin wrapper over packages/research-runner/src/search/arxiv-client.ts. No API key required (arXiv is open). The CLI subcommand backend lands in B-PR1a.
Error contract
On terminal failure (network exhausted), exits non-zero and writes {ok: false, reason} to stdout. Agent continues with remaining providers per §5.5.3.