{"as_of":"2026-08-20T03:24:00Z","caps":{"database_statements":6,"inbound":100,"outbound":100},"context_digest":"sha256:13f10768da1345fad6fc94fae00aa6f788dcd4b8e988b11032a31c8688f0e251","coverage":[{"denominator":15,"lane":"reference_resolution","note":"Typed states for the displayed outbound observations.","records_observed":15,"source":"paper_references, paper_reference_links","source_observed_at":"2026-08-15T23:58:40.082319Z","state":"measured"},{"denominator":29,"lane":"standing_notices","note":"One-hop event checks from named stored sources.","records_observed":29,"source":"scholarly_work_events, retraction_status_cache","source_observed_at":"2026-08-19T06:32:44.657259+00:00","state":"measured"},{"denominator":14,"lane":"inbound_itemization","note":"Pith citing papers itemized under the disclosed page cap.","records_observed":14,"source":"paper_references, paper_reference_links","source_observed_at":"2026-08-15T14:21:42.068204Z","state":"measured"},{"denominator":1,"lane":"external_citation_measurements","note":"A source-named dated measurement, never combined with another source.","records_observed":1,"source":"pith","source_observed_at":"2026-08-05T02:28:24.338817Z","state":"measured"}],"external_citation_measurements":[{"count":0,"observed_at":"2026-08-05T02:28:24.338817Z","source":"pith"}],"inbound":[{"citation":{"cited_paper":{"arxiv_id":"2505.03275","last_updated":"2025-05-06T08:05:35Z","snapshot_observed_at":"2026-08-15T23:53:17.486647Z","submitted_at":"2025-05-06T08:05:35Z","title":"RAG-MCP: Mitigating Prompt Bloat in LLM Tool Selection via Retrieval-Augmented Generation","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2505.03275","snapshot_observed_at":"2026-08-07T13:59:35.395354Z","title":"Rag-mcp: Mitigating prompt bloat in llm tool selection via retrieval-augmented generation.arXiv preprint arXiv:2505.03275, 2025","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2505.20286","last_updated":"2025-05-26T17:58:53Z","snapshot_observed_at":"2026-08-17T12:00:38.078645Z","submitted_at":"2025-05-26T17:58:53Z","title":"Alita: Generalist Agent Enabling Scalable Agentic Reasoning with Minimal Predefinition and Maximal Self-Evolution","version":1},"reference_index":21,"source":"pdf_text","source_observed_at":"2026-08-07T13:59:35.395354Z"},"links":{"cited_paper":"/paper/2505.03275","citing_paper":"/paper/2505.20286"},"observation_digest":"sha256:e2a97fa4a65ea425db3b97b6278c172ec3375d7d5f6fd8a87f5c064c8d3f9cd1","observation_id":"8f61ddac-cbdc-4409-9d26-8e9ec6b2eede","resolution":{"observed_at":"2026-08-07T13:59:35.395354Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2505.03275","last_updated":"2025-05-06T08:05:35Z","snapshot_observed_at":"2026-08-15T23:53:17.486647Z","submitted_at":"2025-05-06T08:05:35Z","title":"RAG-MCP: Mitigating Prompt Bloat in LLM Tool Selection via Retrieval-Augmented Generation","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2505.03275","snapshot_observed_at":"2026-08-07T11:56:25.029907Z","title":"Rag-mcp: Mitigating prompt bloat in llm tool selection via retrieval-augmented genera- tion.arXiv preprint arXiv:2505.03275, 2025","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2506.01056","last_updated":"2025-06-24T06:27:29Z","snapshot_observed_at":"2026-08-18T01:41:19.055056Z","submitted_at":"2025-06-01T15:48:53Z","title":"MCP-Zero: Active Tool Discovery for Autonomous LLM Agents","version":4},"reference_index":7,"source":"pdf_text","source_observed_at":"2026-08-07T11:56:25.029907Z"},"links":{"cited_paper":"/paper/2505.03275","citing_paper":"/paper/2506.01056"},"observation_digest":"sha256:10e0a32c38038aeb6cd036fb558484dfb5cc807d3bc0165503e14d0daf78cd3b","observation_id":"d42f85ba-7247-469c-ac92-86226cee5dd4","resolution":{"observed_at":"2026-08-07T11:56:25.029907Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2505.03275","last_updated":"2025-05-06T08:05:35Z","snapshot_observed_at":"2026-08-15T23:53:17.486647Z","submitted_at":"2025-05-06T08:05:35Z","title":"RAG-MCP: Mitigating Prompt Bloat in LLM Tool Selection via Retrieval-Augmented Generation","version":1},"cited_work":{"arxiv_id":"2505.03275","doi":"10.48550/arxiv.2505.03275","metadata_source":"pith","pith_arxiv_id":"2505.03275","snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"arXiv preprint arXiv:2505.03275 , year=","venue":"cs.AI","work_id":"a17d2b90-a81f-465f-b8f8-558a781a97f0","year":2025},"citing_paper":{"arxiv_id":"2605.10787","last_updated":"2026-05-20T01:39:17Z","snapshot_observed_at":"2026-08-12T22:55:54.365434Z","submitted_at":"2026-05-11T16:20:51Z","title":"ComplexMCP: Evaluation of LLM Agents in Dynamic, Interdependent, and Large-Scale Tool Sandbox","version":1},"reference_index":35,"source":"arxiv_source","source_observed_at":"2026-05-12T04:42:47.496496Z"},"links":{"cited_paper":"/paper/2505.03275","citing_paper":"/paper/2605.10787"},"observation_digest":"sha256:f6832634dd8eda343ec0d31517e9bf4f430056022224434c88f89ed3eb707570","observation_id":"e32a028f-eb2f-4c44-8177-aee07aa0901e","resolution":{"observed_at":"2026-05-12T06:01:23.502690Z","resolver_source":"arxiv_id","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-19T06:32:44.657259+00:00","source":"crossref"},{"observed_at":"2026-07-11T06:20:38.438789+00:00","source":"crossref_status_cache"},{"observed_at":"2026-07-11T06:20:38.438789+00:00","source":"openalex_status_cache"},{"observed_at":"2026-08-19T06:32:39.956319+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2505.03275","last_updated":"2025-05-06T08:05:35Z","snapshot_observed_at":"2026-08-15T23:53:17.486647Z","submitted_at":"2025-05-06T08:05:35Z","title":"RAG-MCP: Mitigating Prompt Bloat in LLM Tool Selection via Retrieval-Augmented Generation","version":1},"cited_work":{"arxiv_id":"2505.03275","doi":"10.48550/arxiv.2505.03275","metadata_source":"pith","pith_arxiv_id":"2505.03275","snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"arXiv preprint arXiv:2505.03275 , year=","venue":"cs.AI","work_id":"a17d2b90-a81f-465f-b8f8-558a781a97f0","year":2025},"citing_paper":{"arxiv_id":"2605.10787","last_updated":"2026-05-20T01:39:17Z","snapshot_observed_at":"2026-08-12T22:55:54.365434Z","submitted_at":"2026-05-11T16:20:51Z","title":"ComplexMCP: Evaluation of LLM Agents in Dynamic, Interdependent, and Large-Scale Tool Sandbox","version":2},"reference_index":35,"source":"arxiv_source","source_observed_at":"2026-05-21T08:08:04.544564Z"},"links":{"cited_paper":"/paper/2505.03275","citing_paper":"/paper/2605.10787"},"observation_digest":"sha256:6e502b1e9835f5daf1f6e1ba483474fc0a370c41d8991e8a0861b7d72404110e","observation_id":"5088c91c-ab89-47fe-9a66-82df53a5aea6","resolution":{"observed_at":"2026-05-21T08:09:51.298758Z","resolver_source":"arxiv_id","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-19T06:32:44.657259+00:00","source":"crossref"},{"observed_at":"2026-07-11T06:20:38.438789+00:00","source":"crossref_status_cache"},{"observed_at":"2026-07-11T06:20:38.438789+00:00","source":"openalex_status_cache"},{"observed_at":"2026-08-19T06:32:39.956319+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2505.03275","last_updated":"2025-05-06T08:05:35Z","snapshot_observed_at":"2026-08-15T23:53:17.486647Z","submitted_at":"2025-05-06T08:05:35Z","title":"RAG-MCP: Mitigating Prompt Bloat in LLM Tool Selection via Retrieval-Augmented Generation","version":1},"cited_work":{"arxiv_id":"2505.03275","doi":"10.48550/arxiv.2505.03275","metadata_source":"pith","pith_arxiv_id":"2505.03275","snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"arXiv preprint arXiv:2505.03275 , year=","venue":"cs.AI","work_id":"a17d2b90-a81f-465f-b8f8-558a781a97f0","year":2025},"citing_paper":{"arxiv_id":"2605.24050","last_updated":"2026-06-23T14:03:22Z","snapshot_observed_at":"2026-08-15T05:30:00.096272Z","submitted_at":"2026-05-21T23:33:31Z","title":"More Skills, Worse Agents? Skill Shadowing Degrades Performance When Expanding Skill Libraries","version":2},"reference_index":1,"source":"pdf_text","source_observed_at":"2026-06-30T15:43:46.616644Z"},"links":{"cited_paper":"/paper/2505.03275","citing_paper":"/paper/2605.24050"},"observation_digest":"sha256:808a762cde7ce8c7523d48847c61269f424bd56521e2bf41ed4a8753b36333e9","observation_id":"8997a0d2-a4f2-4596-a5ae-8a2dad90dbf4","resolution":{"observed_at":"2026-06-30T15:44:48.223839Z","resolver_source":"arxiv_id","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-19T06:32:44.657259+00:00","source":"crossref"},{"observed_at":"2026-07-11T06:20:38.438789+00:00","source":"crossref_status_cache"},{"observed_at":"2026-07-11T06:20:38.438789+00:00","source":"openalex_status_cache"},{"observed_at":"2026-08-19T06:32:39.956319+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2505.03275","last_updated":"2025-05-06T08:05:35Z","snapshot_observed_at":"2026-08-15T23:53:17.486647Z","submitted_at":"2025-05-06T08:05:35Z","title":"RAG-MCP: Mitigating Prompt Bloat in LLM Tool Selection via Retrieval-Augmented Generation","version":1},"cited_work":{"arxiv_id":"2505.03275","doi":"10.48550/arxiv.2505.03275","metadata_source":"pith","pith_arxiv_id":"2505.03275","snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"arXiv preprint arXiv:2505.03275 , year=","venue":"cs.AI","work_id":"a17d2b90-a81f-465f-b8f8-558a781a97f0","year":2025},"citing_paper":{"arxiv_id":"2605.24660","last_updated":"2026-06-07T02:58:55Z","snapshot_observed_at":"2026-08-12T22:54:59.017055Z","submitted_at":"2026-05-23T17:02:56Z","title":"How Many Tools Should an LLM Agent See? A Chance-Corrected Answer","version":2},"reference_index":5,"source":"pdf_text","source_observed_at":"2026-06-30T12:22:59.494989Z"},"links":{"cited_paper":"/paper/2505.03275","citing_paper":"/paper/2605.24660"},"observation_digest":"sha256:2b462922336d4642e64e5221967558ec4a114ebb166610c8db631f5f2458898e","observation_id":"1dac8f22-9f79-4a08-8d3c-9dafa0169d75","resolution":{"observed_at":"2026-06-30T12:24:39.587738Z","resolver_source":"arxiv_id","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-19T06:32:44.657259+00:00","source":"crossref"},{"observed_at":"2026-07-11T06:20:38.438789+00:00","source":"crossref_status_cache"},{"observed_at":"2026-07-11T06:20:38.438789+00:00","source":"openalex_status_cache"},{"observed_at":"2026-08-19T06:32:39.956319+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2505.03275","last_updated":"2025-05-06T08:05:35Z","snapshot_observed_at":"2026-08-15T23:53:17.486647Z","submitted_at":"2025-05-06T08:05:35Z","title":"RAG-MCP: Mitigating Prompt Bloat in LLM Tool Selection via Retrieval-Augmented Generation","version":1},"cited_work":{"arxiv_id":"2505.03275","doi":"10.48550/arxiv.2505.03275","metadata_source":"pith","pith_arxiv_id":"2505.03275","snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"arXiv preprint arXiv:2505.03275 , year=","venue":"cs.AI","work_id":"a17d2b90-a81f-465f-b8f8-558a781a97f0","year":2025},"citing_paper":{"arxiv_id":"2606.06284","last_updated":"2026-06-04T15:24:10Z","snapshot_observed_at":"2026-08-14T07:27:55.488827Z","submitted_at":"2026-06-04T15:24:10Z","title":"ToolChoiceConfusion: Causal Minimal Tool Filtering for Reliable LLM Agents","version":1},"reference_index":3,"source":"pdf_text","source_observed_at":"2026-06-28T01:53:08.249088Z"},"links":{"cited_paper":"/paper/2505.03275","citing_paper":"/paper/2606.06284"},"observation_digest":"sha256:3dfa18b99eeadf66a4fdbcd4f18a829b162c6b821029628bb8a95cef32864532","observation_id":"8b448f87-92ba-41ad-aa33-eb7759ae8afa","resolution":{"observed_at":"2026-07-02T12:46:56.649860Z","resolver_source":"arxiv_id","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-19T06:32:44.657259+00:00","source":"crossref"},{"observed_at":"2026-07-11T06:20:38.438789+00:00","source":"crossref_status_cache"},{"observed_at":"2026-07-11T06:20:38.438789+00:00","source":"openalex_status_cache"},{"observed_at":"2026-08-19T06:32:39.956319+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2505.03275","last_updated":"2025-05-06T08:05:35Z","snapshot_observed_at":"2026-08-15T23:53:17.486647Z","submitted_at":"2025-05-06T08:05:35Z","title":"RAG-MCP: Mitigating Prompt Bloat in LLM Tool Selection via Retrieval-Augmented Generation","version":1},"cited_work":{"arxiv_id":"2505.03275","doi":"10.48550/arxiv.2505.03275","metadata_source":"pith","pith_arxiv_id":"2505.03275","snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"arXiv preprint arXiv:2505.03275 , year=","venue":"cs.AI","work_id":"a17d2b90-a81f-465f-b8f8-558a781a97f0","year":2025},"citing_paper":{"arxiv_id":"2606.17519","last_updated":"2026-06-16T04:55:06Z","snapshot_observed_at":"2026-08-18T06:01:50.633277Z","submitted_at":"2026-06-16T04:55:06Z","title":"Scaling Enterprise Agent Routing: Degradation, Diagnosis, and Recovery","version":1},"reference_index":13,"source":"arxiv_source","source_observed_at":"2026-06-27T01:21:05.583428Z"},"links":{"cited_paper":"/paper/2505.03275","citing_paper":"/paper/2606.17519"},"observation_digest":"sha256:6873c0a09fee36eb16ee9e3c2e5ed00dfdffca5d353fa6ed0eda5a5b1ac17029","observation_id":"25bd6bb9-6a20-4bc3-a3b6-4684fbc9627e","resolution":{"observed_at":"2026-07-03T20:28:55.477253Z","resolver_source":"arxiv_id","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-19T06:32:44.657259+00:00","source":"crossref"},{"observed_at":"2026-07-11T06:20:38.438789+00:00","source":"crossref_status_cache"},{"observed_at":"2026-07-11T06:20:38.438789+00:00","source":"openalex_status_cache"},{"observed_at":"2026-08-19T06:32:39.956319+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2505.03275","last_updated":"2025-05-06T08:05:35Z","snapshot_observed_at":"2026-08-15T23:53:17.486647Z","submitted_at":"2025-05-06T08:05:35Z","title":"RAG-MCP: Mitigating Prompt Bloat in LLM Tool Selection via Retrieval-Augmented Generation","version":1},"cited_work":{"arxiv_id":"2505.03275","doi":"10.48550/arxiv.2505.03275","metadata_source":"pith","pith_arxiv_id":"2505.03275","snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"arXiv preprint arXiv:2505.03275 , year=","venue":"cs.AI","work_id":"a17d2b90-a81f-465f-b8f8-558a781a97f0","year":2025},"citing_paper":{"arxiv_id":"2606.30317","last_updated":"2026-06-29T13:59:41Z","snapshot_observed_at":"2026-07-07T00:04:10.492340Z","submitted_at":"2026-06-29T13:59:41Z","title":"MCP Server Architecture Patterns for LLM-Integrated Applications","version":1},"reference_index":20,"source":"pdf_text","source_observed_at":"2026-06-30T05:18:52.171672Z"},"links":{"cited_paper":"/paper/2505.03275","citing_paper":"/paper/2606.30317"},"observation_digest":"sha256:4c34ecf01478d3ee0dda9aede4b4d1e9ced15dae9dc3eeac9fb2b2f4a9b52816","observation_id":"09375685-26c7-454a-848d-4fac43b7ac01","resolution":{"observed_at":"2026-06-30T14:44:45.894718Z","resolver_source":"arxiv_id","status":"metadata_mismatch"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-19T06:32:44.657259+00:00","source":"crossref"},{"observed_at":"2026-07-11T06:20:38.438789+00:00","source":"crossref_status_cache"},{"observed_at":"2026-07-11T06:20:38.438789+00:00","source":"openalex_status_cache"},{"observed_at":"2026-08-19T06:32:39.956319+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2505.03275","last_updated":"2025-05-06T08:05:35Z","snapshot_observed_at":"2026-08-15T23:53:17.486647Z","submitted_at":"2025-05-06T08:05:35Z","title":"RAG-MCP: Mitigating Prompt Bloat in LLM Tool Selection via Retrieval-Augmented Generation","version":1},"cited_work":{"arxiv_id":"2505.03275","doi":"10.48550/arxiv.2505.03275","metadata_source":"pith","pith_arxiv_id":"2505.03275","snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"arXiv preprint arXiv:2505.03275 , year=","venue":"cs.AI","work_id":"a17d2b90-a81f-465f-b8f8-558a781a97f0","year":2025},"citing_paper":{"arxiv_id":"2607.05936","last_updated":"2026-07-07T07:38:28Z","snapshot_observed_at":"2026-08-17T00:21:17.412530Z","submitted_at":"2026-07-07T07:38:28Z","title":"Mitigating Errors in LLM-Generated Web API Invocations via Retrieval-Augmented Generation and Constrained Decoding","version":1},"reference_index":12,"source":"pdf_text","source_observed_at":"2026-07-08T20:32:18.337262Z"},"links":{"cited_paper":"/paper/2505.03275","citing_paper":"/paper/2607.05936"},"observation_digest":"sha256:6ccbc0ece06cb7c744b88064fae7e2f32ea1ef8b4b2554d7e5a52b8b846824d0","observation_id":"c89f0f96-942f-4eaf-bec9-a7e30c65e3b6","resolution":{"observed_at":"2026-07-08T20:35:34.146738Z","resolver_source":"local_arxiv","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-19T06:32:44.657259+00:00","source":"crossref"},{"observed_at":"2026-07-11T06:20:38.438789+00:00","source":"crossref_status_cache"},{"observed_at":"2026-07-11T06:20:38.438789+00:00","source":"openalex_status_cache"},{"observed_at":"2026-08-19T06:32:39.956319+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2505.03275","last_updated":"2025-05-06T08:05:35Z","snapshot_observed_at":"2026-08-15T23:53:17.486647Z","submitted_at":"2025-05-06T08:05:35Z","title":"RAG-MCP: Mitigating Prompt Bloat in LLM Tool Selection via Retrieval-Augmented Generation","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2505.03275","snapshot_observed_at":"2026-08-01T22:57:35.992982Z","title":null,"venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.15593","last_updated":"2026-07-17T03:40:49Z","snapshot_observed_at":"2026-08-18T00:58:45.406033Z","submitted_at":"2026-07-17T03:40:49Z","title":"Scalable LLM Agent Tool Access in the Cloud","version":1},"reference_index":20,"source":"pdf_text","source_observed_at":"2026-08-01T22:57:35.992982Z"},"links":{"cited_paper":"/paper/2505.03275","citing_paper":"/paper/2607.15593"},"observation_digest":"sha256:52480264085ed192525093e78767defe92a73bdf3bc6e16efef7bec762bbdacd","observation_id":"4badcbb6-f770-47e4-869e-881915469cef","resolution":{"observed_at":"2026-08-01T22:57:35.992982Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2505.03275","last_updated":"2025-05-06T08:05:35Z","snapshot_observed_at":"2026-08-15T23:53:17.486647Z","submitted_at":"2025-05-06T08:05:35Z","title":"RAG-MCP: Mitigating Prompt Bloat in LLM Tool Selection via Retrieval-Augmented Generation","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2505.03275","snapshot_observed_at":"2026-08-01T18:19:37.745270Z","title":null,"venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.17347","last_updated":"2026-07-19T17:17:47Z","snapshot_observed_at":"2026-08-18T16:41:34.854623Z","submitted_at":"2026-07-19T17:17:47Z","title":"Adapting Embedding Models for Agent Capability Retrieval","version":1},"reference_index":2,"source":"pdf_text","source_observed_at":"2026-08-01T18:19:37.745270Z"},"links":{"cited_paper":"/paper/2505.03275","citing_paper":"/paper/2607.17347"},"observation_digest":"sha256:bc791e669ac876bc2af3b5d9f8f7f29720dffc8420ba4d1894fbeba0c28cde76","observation_id":"241c025d-ad15-4526-85c8-83fbe3e0bbf7","resolution":{"observed_at":"2026-08-01T18:19:37.745270Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2505.03275","last_updated":"2025-05-06T08:05:35Z","snapshot_observed_at":"2026-08-15T23:53:17.486647Z","submitted_at":"2025-05-06T08:05:35Z","title":"RAG-MCP: Mitigating Prompt Bloat in LLM Tool Selection via Retrieval-Augmented Generation","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2505.03275","snapshot_observed_at":"2026-08-06T00:39:14.332210Z","title":"Gan and Q","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2608.01050","last_updated":"2026-08-02T07:32:49Z","snapshot_observed_at":"2026-08-17T18:02:26.052022Z","submitted_at":"2026-08-02T07:32:49Z","title":"Don't Offer What Can't Be Done: Deterministic Executability Gating for LLM Skill Selection at Scale","version":1},"reference_index":9,"source":"pdf_text","source_observed_at":"2026-08-06T00:39:14.332210Z"},"links":{"cited_paper":"/paper/2505.03275","citing_paper":"/paper/2608.01050"},"observation_digest":"sha256:912509dd47b2db20770fca1212b99ee0a9cf71b4e60f3e56ee8c7cc7e71d77dc","observation_id":"6037c6c8-ef73-4d6d-8125-c4b49e369beb","resolution":{"observed_at":"2026-08-06T00:39:14.332210Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2505.03275","last_updated":"2025-05-06T08:05:35Z","snapshot_observed_at":"2026-08-15T23:53:17.486647Z","submitted_at":"2025-05-06T08:05:35Z","title":"RAG-MCP: Mitigating Prompt Bloat in LLM Tool Selection via Retrieval-Augmented Generation","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2505.03275","snapshot_observed_at":"2026-08-15T14:21:42.068204Z","title":null,"venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2608.10530","last_updated":"2026-08-11T06:11:26Z","snapshot_observed_at":"2026-08-19T18:52:42.862550Z","submitted_at":"2026-08-11T06:11:26Z","title":"On Understanding, Identifying, and Mitigating Vulnerabilities in Agentic Large Language Models","version":1},"reference_index":48,"source":"pdf_text","source_observed_at":"2026-08-15T14:21:42.068204Z"},"links":{"cited_paper":"/paper/2505.03275","citing_paper":"/paper/2608.10530"},"observation_digest":"sha256:a11257f39e6b58c00b8d35f68bc80521d21e0e0e3467213d852a9f0a225b908b","observation_id":"a785bded-1e3d-49cf-9003-73025b12f846","resolution":{"observed_at":"2026-08-15T14:21:42.068204Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}}],"links":{"evidence":"/evidence","html":"/paper/2505.03275/citation-record","integrity":"/paper/2505.03275/integrity","json":"/paper/2505.03275/citation-record.json","paper":"/paper/2505.03275"},"outbound":[{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-15T23:58:40.328666Z","title":"anthropic.com/news/model-context-protocol","venue":null,"work_id":"1735c690-0d26-49e7-bcb4-962af37581e3","year":2024},"citing_paper":{"arxiv_id":"2505.03275","last_updated":"2025-05-06T08:05:35Z","snapshot_observed_at":"2026-08-15T23:53:17.486647Z","submitted_at":"2025-05-06T08:05:35Z","title":"RAG-MCP: Mitigating Prompt Bloat in LLM Tool Selection via Retrieval-Augmented Generation","version":1},"reference_index":1,"source":"pdf_text","source_observed_at":"2026-08-15T23:58:40.025920Z"},"links":{"citing_paper":"/paper/2505.03275"},"observation_digest":"sha256:14ea4565c81da37851092c47b98cd907a3b7ab8dcd047fe6969ed782e221f196","observation_id":"58a95336-9246-45a7-9fbe-f8f457d718af","resolution":{"observed_at":"2026-08-15T23:58:40.333064Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-19T06:32:44.657259+00:00","source":"crossref"},{"observed_at":"2026-08-19T06:32:39.956319+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-15T23:58:40.316916Z","title":"nvidia.com/blog/what-is-retrieval-augmented-generation/","venue":null,"work_id":"89b7b244-45ea-413e-b251-ca33035e8cdf","year":2025},"citing_paper":{"arxiv_id":"2505.03275","last_updated":"2025-05-06T08:05:35Z","snapshot_observed_at":"2026-08-15T23:53:17.486647Z","submitted_at":"2025-05-06T08:05:35Z","title":"RAG-MCP: Mitigating Prompt Bloat in LLM Tool Selection via Retrieval-Augmented Generation","version":1},"reference_index":2,"source":"pdf_text","source_observed_at":"2026-08-15T23:58:40.030336Z"},"links":{"citing_paper":"/paper/2505.03275"},"observation_digest":"sha256:402c0fc240ff7f9d48fff192dea33fbb9dcf759b08e0d37b9a113d53253bd20f","observation_id":"6ff10936-6ea4-4480-9069-01517b87ccc2","resolution":{"observed_at":"2026-08-15T23:58:40.321139Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-19T06:32:44.657259+00:00","source":"crossref"},{"observed_at":"2026-08-19T06:32:39.956319+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2412.01130","last_updated":"2024-12-04T03:34:42Z","snapshot_observed_at":"2026-08-19T16:48:32.754333Z","submitted_at":"2024-12-02T05:10:41Z","title":"Enhancing Function-Calling Capabilities in LLMs: Strategies for Prompt Formats, Data Integration, and Multilingual Translation","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2412.01130","snapshot_observed_at":"2026-08-15T23:58:40.033980Z","title":"arXiv preprint arXiv:2412.01130 (2024)","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2505.03275","last_updated":"2025-05-06T08:05:35Z","snapshot_observed_at":"2026-08-15T23:53:17.486647Z","submitted_at":"2025-05-06T08:05:35Z","title":"RAG-MCP: Mitigating Prompt Bloat in LLM Tool Selection via Retrieval-Augmented Generation","version":1},"reference_index":3,"source":"pdf_text","source_observed_at":"2026-08-15T23:58:40.033980Z"},"links":{"cited_paper":"/paper/2412.01130","citing_paper":"/paper/2505.03275"},"observation_digest":"sha256:4d59101d1c9333fb4b61e85c8e4475297d3963ec5affae3c7fa2b04bd6e04dd9","observation_id":"ef03cc61-b63e-4705-87eb-170870b54eab","resolution":{"observed_at":"2026-08-15T23:58:40.033980Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2312.10997","last_updated":"2024-03-27T09:16:57Z","snapshot_observed_at":"2026-07-06T02:11:23.670680Z","submitted_at":"2023-12-18T07:47:33Z","title":"Retrieval-Augmented Generation for Large Language Models: A Survey","version":5},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2312.10997","snapshot_observed_at":"2026-08-15T23:58:40.038019Z","title":"arXiv preprint arXiv:2312.109972 (2023) 12 Gan and Sun","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2505.03275","last_updated":"2025-05-06T08:05:35Z","snapshot_observed_at":"2026-08-15T23:53:17.486647Z","submitted_at":"2025-05-06T08:05:35Z","title":"RAG-MCP: Mitigating Prompt Bloat in LLM Tool Selection via Retrieval-Augmented Generation","version":1},"reference_index":4,"source":"pdf_text","source_observed_at":"2026-08-15T23:58:40.038019Z"},"links":{"cited_paper":"/paper/2312.10997","citing_paper":"/paper/2505.03275"},"observation_digest":"sha256:fe2a237dcb47e6bc7169f6d0c85a225fa0a66f3f61c1f9dd7e473473829be882","observation_id":"9ef93b14-a11b-4428-83c2-73ec5f053583","resolution":{"observed_at":"2026-08-15T23:58:40.038019Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-15T23:58:40.304639Z","title":null,"venue":null,"work_id":"afd60666-c3ae-4072-baa6-e8b26ae45b9d","year":2024},"citing_paper":{"arxiv_id":"2505.03275","last_updated":"2025-05-06T08:05:35Z","snapshot_observed_at":"2026-08-15T23:53:17.486647Z","submitted_at":"2025-05-06T08:05:35Z","title":"RAG-MCP: Mitigating Prompt Bloat in LLM Tool Selection via Retrieval-Augmented Generation","version":1},"reference_index":5,"source":"pdf_text","source_observed_at":"2026-08-15T23:58:40.042218Z"},"links":{"citing_paper":"/paper/2505.03275"},"observation_digest":"sha256:d21a66689ec80afe275ec18474389b522ac53342f9c2eea9de98ddf6c1859ecd","observation_id":"134c826b-9c05-4537-9df5-941ffb660106","resolution":{"observed_at":"2026-08-15T23:58:40.308343Z","resolver_source":"raw_fallback","status":"unresolved"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-19T06:32:44.657259+00:00","source":"crossref"},{"observed_at":"2026-08-19T06:32:39.956319+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-15T23:58:40.293338Z","title":"In: Larochelle, H., Ranzato, M., Hadsell, R., Balcan, M., Lin, H","venue":null,"work_id":"f296d25c-0e1c-4061-b41e-dd586ce964cc","year":2020},"citing_paper":{"arxiv_id":"2505.03275","last_updated":"2025-05-06T08:05:35Z","snapshot_observed_at":"2026-08-15T23:53:17.486647Z","submitted_at":"2025-05-06T08:05:35Z","title":"RAG-MCP: Mitigating Prompt Bloat in LLM Tool Selection via Retrieval-Augmented Generation","version":1},"reference_index":6,"source":"pdf_text","source_observed_at":"2026-08-15T23:58:40.046520Z"},"links":{"citing_paper":"/paper/2505.03275"},"observation_digest":"sha256:6317fe0f73c5172a9f78042854a6313d2e175e47bb217f74989d41698cb96e96","observation_id":"06575350-2197-4a95-8d82-2366acf3eb64","resolution":{"observed_at":"2026-08-15T23:58:40.297315Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-19T06:32:44.657259+00:00","source":"crossref"},{"observed_at":"2026-08-19T06:32:39.956319+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2412.19437","last_updated":"2025-02-18T17:26:38Z","snapshot_observed_at":"2026-08-18T18:18:37.449517Z","submitted_at":"2024-12-27T04:03:16Z","title":"DeepSeek-V3 Technical Report","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2412.19437","snapshot_observed_at":"2026-08-15T23:58:40.050943Z","title":"arXiv preprint arXiv:2412.19437 (2024)","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2505.03275","last_updated":"2025-05-06T08:05:35Z","snapshot_observed_at":"2026-08-15T23:53:17.486647Z","submitted_at":"2025-05-06T08:05:35Z","title":"RAG-MCP: Mitigating Prompt Bloat in LLM Tool Selection via Retrieval-Augmented Generation","version":1},"reference_index":7,"source":"pdf_text","source_observed_at":"2026-08-15T23:58:40.050943Z"},"links":{"cited_paper":"/paper/2412.19437","citing_paper":"/paper/2505.03275"},"observation_digest":"sha256:e27d20036ae87e2ffac006fb18d4091dd45723132411726e8f15f52e866a7b0b","observation_id":"eb7d2e07-69c5-4734-a59a-1baa66fa9be4","resolution":{"observed_at":"2026-08-15T23:58:40.050943Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2504.11094","last_updated":"2025-04-18T10:39:23Z","snapshot_observed_at":"2026-08-16T12:40:56.136470Z","submitted_at":"2025-04-15T11:40:12Z","title":"Evaluation Report on MCP Servers","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2504.11094","snapshot_observed_at":"2026-08-15T23:58:40.055045Z","title":null,"venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2505.03275","last_updated":"2025-05-06T08:05:35Z","snapshot_observed_at":"2026-08-15T23:53:17.486647Z","submitted_at":"2025-05-06T08:05:35Z","title":"RAG-MCP: Mitigating Prompt Bloat in LLM Tool Selection via Retrieval-Augmented Generation","version":1},"reference_index":8,"source":"pdf_text","source_observed_at":"2026-08-15T23:58:40.055045Z"},"links":{"cited_paper":"/paper/2504.11094","citing_paper":"/paper/2505.03275"},"observation_digest":"sha256:c09dce241bd31018f64e6c24c69024b038ad3dfdd2e48d26828045d9114a8f1d","observation_id":"0d94cc82-81d0-4207-80a5-fa5ef109cbb5","resolution":{"observed_at":"2026-08-15T23:58:40.055045Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-15T23:58:40.282083Z","title":null,"venue":null,"work_id":"4f9c3559-92f4-4232-a5ca-a3e3afe80866","year":2022},"citing_paper":{"arxiv_id":"2505.03275","last_updated":"2025-05-06T08:05:35Z","snapshot_observed_at":"2026-08-15T23:53:17.486647Z","submitted_at":"2025-05-06T08:05:35Z","title":"RAG-MCP: Mitigating Prompt Bloat in LLM Tool Selection via Retrieval-Augmented Generation","version":1},"reference_index":9,"source":"pdf_text","source_observed_at":"2026-08-15T23:58:40.058936Z"},"links":{"citing_paper":"/paper/2505.03275"},"observation_digest":"sha256:b30947294e5dfb1c567473f0dc57ca2b0a03151599112bbf4d02ff854e8092d3","observation_id":"6d624ff9-7338-4900-b0e7-336f97b2fc32","resolution":{"observed_at":"2026-08-15T23:58:40.285911Z","resolver_source":"raw_fallback","status":"unresolved"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-19T06:32:44.657259+00:00","source":"crossref"},{"observed_at":"2026-08-19T06:32:39.956319+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-15T23:58:40.269736Z","title":null,"venue":null,"work_id":"96ba1cae-0bc5-45f8-8e15-63002b4d098b","year":null},"citing_paper":{"arxiv_id":"2505.03275","last_updated":"2025-05-06T08:05:35Z","snapshot_observed_at":"2026-08-15T23:53:17.486647Z","submitted_at":"2025-05-06T08:05:35Z","title":"RAG-MCP: Mitigating Prompt Bloat in LLM Tool Selection via Retrieval-Augmented Generation","version":1},"reference_index":10,"source":"pdf_text","source_observed_at":"2026-08-15T23:58:40.063010Z"},"links":{"citing_paper":"/paper/2505.03275"},"observation_digest":"sha256:55e82961608cb0ff70e6634a66dd1e45caf71eb8c6ab79acaeaf565c1aa96523","observation_id":"520a8c9e-78aa-4347-bd43-39b4cec84688","resolution":{"observed_at":"2026-08-15T23:58:40.273749Z","resolver_source":"raw_fallback","status":"unresolved"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-19T06:32:44.657259+00:00","source":"crossref"},{"observed_at":"2026-08-19T06:32:39.956319+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-15T23:58:40.257832Z","title":null,"venue":null,"work_id":"2be293b7-255d-4430-8ed9-d33e5debdf4e","year":2023},"citing_paper":{"arxiv_id":"2505.03275","last_updated":"2025-05-06T08:05:35Z","snapshot_observed_at":"2026-08-15T23:53:17.486647Z","submitted_at":"2025-05-06T08:05:35Z","title":"RAG-MCP: Mitigating Prompt Bloat in LLM Tool Selection via Retrieval-Augmented Generation","version":1},"reference_index":11,"source":"pdf_text","source_observed_at":"2026-08-15T23:58:40.067087Z"},"links":{"citing_paper":"/paper/2505.03275"},"observation_digest":"sha256:abbbd94605d1f9dc01a8344d3d64082206ad661b35b76af4cd334d3bdf045e78","observation_id":"62e91ba0-dd3e-489a-87af-b4aeeb33295c","resolution":{"observed_at":"2026-08-15T23:58:40.261888Z","resolver_source":"raw_fallback","status":"unresolved"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-19T06:32:44.657259+00:00","source":"crossref"},{"observed_at":"2026-08-19T06:32:39.956319+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-15T23:58:40.070644Z","title":"Advances in Neural Information Processing Systems 37, 126544–126565 (2024)","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2505.03275","last_updated":"2025-05-06T08:05:35Z","snapshot_observed_at":"2026-08-15T23:53:17.486647Z","submitted_at":"2025-05-06T08:05:35Z","title":"RAG-MCP: Mitigating Prompt Bloat in LLM Tool Selection via Retrieval-Augmented Generation","version":1},"reference_index":12,"source":"pdf_text","source_observed_at":"2026-08-15T23:58:40.070644Z"},"links":{"citing_paper":"/paper/2505.03275"},"observation_digest":"sha256:67a64791d6c4a68740325e6efe993213224cd3d17561715b2526bbcec785aa7d","observation_id":"e45e7e9d-c386-4a31-bdaf-d9668955ab2e","resolution":{"observed_at":"2026-08-15T23:58:40.070644Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-15T23:58:40.239325Z","title":"In: Oh, A., Naumann, T., Glober- son, A., Saenko, K., Hardt, M., Levine, S","venue":null,"work_id":"3cbab23c-a3e1-4f31-9a9d-295b8e35bfe3","year":2023},"citing_paper":{"arxiv_id":"2505.03275","last_updated":"2025-05-06T08:05:35Z","snapshot_observed_at":"2026-08-15T23:53:17.486647Z","submitted_at":"2025-05-06T08:05:35Z","title":"RAG-MCP: Mitigating Prompt Bloat in LLM Tool Selection via Retrieval-Augmented Generation","version":1},"reference_index":13,"source":"pdf_text","source_observed_at":"2026-08-15T23:58:40.074574Z"},"links":{"citing_paper":"/paper/2505.03275"},"observation_digest":"sha256:906bbf8c8f1e10d60b062fa29771cc9bc39399bffabb395c5c7ae91235d2efa0","observation_id":"d0a833ab-a540-4240-a8b5-a9ab516177a8","resolution":{"observed_at":"2026-08-15T23:58:40.243713Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-19T06:32:44.657259+00:00","source":"crossref"},{"observed_at":"2026-08-19T06:32:39.956319+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-15T23:58:40.212609Z","title":null,"venue":null,"work_id":"4f391387-f3e3-4e2d-805d-d9a9ea15f845","year":2025},"citing_paper":{"arxiv_id":"2505.03275","last_updated":"2025-05-06T08:05:35Z","snapshot_observed_at":"2026-08-15T23:53:17.486647Z","submitted_at":"2025-05-06T08:05:35Z","title":"RAG-MCP: Mitigating Prompt Bloat in LLM Tool Selection via Retrieval-Augmented Generation","version":1},"reference_index":14,"source":"pdf_text","source_observed_at":"2026-08-15T23:58:40.078569Z"},"links":{"citing_paper":"/paper/2505.03275"},"observation_digest":"sha256:d6f915ccb520cc8e417835ca9f7960afe97f8bf18f35e8beaa613a4453029512","observation_id":"b1cabb30-0222-4175-9be9-54c8294e6b21","resolution":{"observed_at":"2026-08-15T23:58:40.231410Z","resolver_source":"raw_fallback","status":"unresolved"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-19T06:32:44.657259+00:00","source":"crossref"},{"observed_at":"2026-08-19T06:32:39.956319+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-15T23:58:40.157881Z","title":"Evaluation Report on MCP Servers","venue":null,"work_id":"7d94f20b-d14f-4bf4-9640-1e8690eaaf61","year":2023},"citing_paper":{"arxiv_id":"2505.03275","last_updated":"2025-05-06T08:05:35Z","snapshot_observed_at":"2026-08-15T23:53:17.486647Z","submitted_at":"2025-05-06T08:05:35Z","title":"RAG-MCP: Mitigating Prompt Bloat in LLM Tool Selection via Retrieval-Augmented Generation","version":1},"reference_index":15,"source":"pdf_text","source_observed_at":"2026-08-15T23:58:40.082319Z"},"links":{"citing_paper":"/paper/2505.03275"},"observation_digest":"sha256:284780aadc0c751037cb5b56dca3a43ccd3e1981c3dd291a53ec03fd75228b98","observation_id":"da08390f-1d8a-4a31-a54b-1f345f231894","resolution":{"observed_at":"2026-08-15T23:58:40.189345Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-19T06:32:44.657259+00:00","source":"crossref"},{"observed_at":"2026-08-19T06:32:39.956319+00:00","source":"retraction_watch"}],"state":"measured"}}],"paper":{"arxiv_id":"2505.03275","last_updated":"2025-05-06T08:05:35Z","latest_version":1,"primary_category":"cs.AI","snapshot_observed_at":"2026-08-15T23:53:17.486647Z","submitted_at":"2025-05-06T08:05:35Z","title":"RAG-MCP: Mitigating Prompt Bloat in LLM Tool Selection via Retrieval-Augmented Generation"},"reference_resolution":{"displayed":15,"state_counts":{"malformed_identifier":0,"metadata_mismatch":0,"parse_uncertain":0,"unresolved":10,"verified_exact":0,"verified_fuzzy":5},"total_outbound_references":15},"refusal":"A citation records a reference. It does not transfer a finding from one paper to another.","schema":"pith.paper-citation-record.v1","standing_sources":[{"observed_at":"2026-08-19T06:32:44.657259+00:00","source":"crossref"},{"observed_at":"2026-08-19T06:32:39.956319+00:00","source":"retraction_watch"}],"thesis":"As of 20 August 2026, this Paper Citation Record lists 15 of 15 outbound references and 14 inbound Pith citation observations for arXiv:2505.03275."}