openevidence-rate-limits

Triage OpenEvidence access, latency, or capacity symptoms without asserting undocumented quotas or retry contracts. Use when working with OpenEvidence in a healthcare organization. Trigger with "openevidence rate limits", "OpenEvidence capacity", or a matching workflow request.

Allowed Tools

ReadGlobGrepWebFetchWriteEdit

Provided by Plugin

openevidence-pack

Operator-grade OpenEvidence pack with 24 source-grounded skills for clinical evidence review, safe adoption, privacy, security, and product workflow governance.

saas packs v1.14.0
View Plugin

Installation

This skill is included in the openevidence-pack plugin:

/plugin install openevidence-pack@claude-code-plugins-plus

Click to copy

Instructions

OpenEvidence Access and Capacity Triage

Overview

Distinguish local connectivity, account/feature access, service behavior, and clinical urgency using safe evidence. Keep inputs minimal, separate observed facts from assumptions, and leave consequential decisions with the named accountable owner.

Prerequisites

  • A clearly bounded workflow, accountable clinical owner, and organizational policy
  • Current first-party OpenEvidence documentation and applicable institution agreements
  • Synthetic or properly authorized minimum-necessary data

Tool Discipline

Use Read, Glob, and Grep to inspect supplied policies, plans, and evidence. Use WebFetch only for current first-party OpenEvidence documentation. Use Write or Edit only when the user requests a named deliverable with an approved destination. Never expose credentials, PHI, recordings, or unrestricted environment output.

Current Contract

  • No public numeric request limit or API retry contract was found for OpenEvidence.
  • Browser/app symptoms must not be translated into invented HTTP limits.
  • Urgent care needs must use an independent clinical fallback.

Authentication

Use only the official OpenEvidence web/mobile sign-in or an institution-approved access path. Do not invent API keys, OAuth clients, SDK credentials, service accounts, or private endpoints. Never ask a user to reveal a password, session token, cookie, or recovery code.

Instructions

  1. Record UTC time, surface, feature, user cohort, duration, impact, and whether patient care is time-sensitive.
  2. Move urgent work to the approved clinical fallback before diagnosis.
  3. Check local network/browser/device conditions and current first-party status/support information.
  4. Reproduce once with synthetic content if safe; do not hammer the service or probe private endpoints.
  5. Compare affected users/surfaces and capture a redacted support packet.
  6. Report observed capacity symptoms, not a guessed quota, with workaround, owner, and escalation state.

Approval Boundaries

Do not create or share accounts; change access, roles, agreements, consent, retention, or security settings; enter PHI; record a conversation; copy content into another system; contact a patient; make a diagnosis or treatment decision; submit billing; transmit a support packet; run a production pilot; or represent vendor capabilities without explicit approval from the accountable owner. A qualified professional remains responsible for clinical decisions.

Output

Return scope, current first-party evidence and date, data classification, workflow or findings, citations reviewed, assumptions rejected, clinical and governance owners, approval state, unresolved risk, and the exact next action. Redact patient and credential data.

Error Handling

Condition Response
Repeated failures Stop retries and use the approved fallback.
Only one account affected Check authorized account/feature access without credential sharing.
No published limit Say unknown and ask support or the contract owner for written terms.

Examples

This compact example shows the minimum reviewable handoff; adapt fields to the approved workflow without adding sensitive data.

Input:


symptom=slow Ask response; surface=web; duration=20m; urgency=non-urgent

Expected handoff:


scope=3 users; local-network=cleared; quota=unknown; fallback=active; support=opened

Resources

Ready to use openevidence-pack?