Agent skill · software engineering · jeremylongshore
abridge-rate-limits
Implement Abridge rate limiting, backoff, and session throttling patterns.
Why this skill is useful
Adds executable scripts for managing Abridge API rate limits and session throttling that are not publicly documented.
What it needs
Requires npm installed locally. About 4k tokens when loaded. Last updated 2026-08-07. 2,604 stars on the source repository.
What this skill does
Abridge Rate Limits Overview Abridge enforces rate limits per organization to ensure platform stability across thousands of concurrent clinical encounters. Limits vary by contract tier and are enforced at the orgid level. Rate Limit Tiers Tier Concurrent Sessions API Calls/min Note Generations/min WebSocket Connections ------ --------------------- --------------- --------------------- ----------------------- Pilot 10 60 10 10 Standard 100 600 50 100 Enterprise 1,000+ 6,000 500 1,000 Response Headers Prerequisites Confirm the organization’s contracted limits with the Abridge account owner; do not treat the illustrative tiers below as an entitlement. Configure an authenticated sandbox client and observability that records aggregate counts, latency, and retry outcomes without retaining PHI. Define a patient-care-safe fallback for requests that cannot be completed within the retry budget. Instructions Step 1: Rate Limit Aware Client Step 2: Concurrent Session Manager Step 3: 429 Error Handler with Retry-After Step 4: Usage Monitoring Output Rate-limit-aware API client with header parsing Concurrent session pool with Bottleneck throttling Automatic 429 retry with Retry-After header support Usage monitoring with utilization alerts Examples In a sandbox load test, configure the session limiter below the documented organization cap and deliberately send enough synthetic requests to receive a 429. Verify that the client honors Retry-After, queues new work, and emits an aggregate retry metric without logging request payloads. When the retry budget is exhausted, return a controlled unavailable result to the caller and route the encounter to the approved manual fallback; do not retry indefinitely or discard a clinically important request without an operator-visible signal. …
How to use it
Reference it in AdaL, Claude Code, Cursor or any coding agent — nothing to install:
@skills jeremylongshore/abridge-rate-limits