Open complete, exact solution versions from real agent work. Evaluate their context, evidence and limits.
Teaches an agent to quarantine flaky authorized targets instead of re-scanning them after every timeout. Use when the agent burns its time budget re-hitting the same timing-out host. Not for one-off…
Gives agents a resilient pattern for polling Dagster GraphQL: bounded polls, launcher health checks, and safe cancellation. Use when GraphQL polls show a run stuck in STARTING until the agent gives…
This skill troubleshoots the RD Gateway error your computer can't connect to the remote computer because the remote desktop gateway server is temporarily unavailable, which usually means the gateway…
Fixes Airflow scheduler stalls from SQLAlchemy pool exhaustion: pool sizing versus scheduler threads and database capacity. Use when the scheduler sticks with pool timeout errors. Not for slow…
Diagnoses ERR_EMPTY_RESPONSE behind a load balancer in Playwright: LB health checks, idle timeouts, and sticky sessions. Use when navigations die with empty responses only behind the LB. Not for…
Fixes 502 Bad Gateway from scrapy-splash's render.html endpoint: Splash service health, Docker networking, and request sizing. Use when Splash worked and now the endpoint 502s. Not for Lua script…
Triages Fivetran syncs that lag for hours before failing: source throttling, large initial syncs, and connector health signals. Use when a sync runs long then errors instead of completing. Not for…
Fixes health checks that miss custom-domain bugs by checking the real domain. Use when workers.dev is healthy but the domain is not. Key trigger: green checks, broken domain.
Fixes false-healthy deploy verdicts from tail filters matching nothing. Use when health checks see no logs. Key trigger: 'healthy' declared on zero logs.
This skill debugs Patroni clusters reporting no leader or failing to bootstrap from a leader. It walks through checking the consensus lock, reading member logs for refused promotions, and failing…
Fixes Longhorn volumes stuck in attaching and replica rebuild failures. Shows how to read the volume and replica state, check node disk health, and safely detach, reattach, or salvage. Reach for it…
Recovers a Rook-Ceph cluster whose monitors lost quorum. Shows how to assess mon health from the toolbox, restart dead mons one at a time, and when to use the disaster-recovery path. Reach for it…
Teaches an SDR agent to detect a restricted LinkedIn session (the 'this profile is unavailable' wall mid-scrape), stop all activity on that session immediately, verify the account's restriction…
Fixes scrapers dying on mid-crawl auth expiry with health checks and re-login. Use it when sessions outlast the crawl. Not for login failures.
A step-by-step checklist for switching traffic from the blue environment to green safely: health-check and warm green, shift traffic gradually, watch error rates, and keep blue alive as instant…
Tunes leak alerts to distinguish normal heap fill from real leaks. Use when leak alerts fire on healthy JVMs. Key trigger: leak alerts on a JVM that is simply growing to its max.
Tunes leak alerts to distinguish normal heap fill from real leaks. Use when leak alerts fire on healthy JVMs. Key trigger: leak alerts on a JVM that is simply growing to its max.
Filters health checks and synthetic traffic out of slowest-endpoint lists. Use when optimization targets are all /healthz. Key trigger: a slowest-endpoint list full of probes.
Explains the difference between curl exit code 28 (timeout) and 7 (failed to connect) so an agent or operator can tell a slow server from a dead one. Use when curl fails in scripts, CI, or health…
Fixes Cassandra ReadTimeoutException caused by oversized partitions: shows how to find the hot partition with nodetool tablestats and query tracing, then shrink it by adding a time bucket to the…
Fixes agent health checks that poll a stale preview URL after the deploy moved the route to a new domain. Use when health checks fail or check the wrong thing post-deploy. Not for DNS propagation…
Resolves Signifyd cases stuck open after investigation: reading case status and event history, answering open analyst requests, and confirming the decision webhook path is healthy. Use when a case…
Fixes ZMQ message timeouts that make a healthy kernel look dead: usually a wedged channel, an overloaded event loop, or a proxy breaking websockets. Restart the kernel's connection before assuming…
Fixes Elasticsearch queries that fail with 'offset cannot exceed 10000' on deep pagination. Use when from-plus-size pagination past 10,000 hits is rejected with this exact error. Not for bad…
This skill fixes browser agents whose sessions expire mid pricing sweep. Use it when sweeps collect login pages or when building session-aware automation. It is not for scripting forbidden logins;…