Menu
When two agents produce overlapping skills in one session, consolidate by merging unique content into the highest-confidence existing skill.
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
Based on SOC occupation classification
Pattern for keeping a rolling temperature-trend sample buffer in Hubitat state, computing slope over a configurable window, and classifying rising/falling/steady/unknown.
Implement Tuya Local v3.3 Groovy drivers on Hubitat using rawSocket, AES-128-ECB, queued retries, and defensive frame parsing.
Use Hubitat async HTTP plus a request queue to authenticate with Cognito, cache tokens in state, refresh proactively, and replay a single 401-failed request.
Never let cached-state dedup bypass session validity in cloud-backed Hubitat drivers.
Standard guard pattern for async HTTP response callbacks in Hubitat drivers
Correct formula for converting lat/lng degree differences to miles, with pole clamp for numerical stability
| name | skill-consolidation |
| description | When two agents produce overlapping skills in one session, consolidate by merging unique content into the highest-confidence existing skill. |
| domain | team-process |
| confidence | low |
| source | earned |
When multiple agents independently produce skills on the same topic in a single session, consolidate rather than keeping duplicates:
git rm.Consolidation reduces skill fragmentation, keeps the skill graph navigable, and preserves the confidence lifecycle (low → medium/high only after repeated independent confirmation).