Commit Graph

806 Commits

Author SHA1 Message Date
Owen f7aca85417 link the usage with the session to display together 2026-08-11 16:15:01 -04:00
miloschwartz c42df737b0 improve model provider selection algorithm 2026-08-11 16:10:46 -04:00
Owen bc2f291352 normalize the requests to also store in the log for viewing later 2026-08-11 14:45:25 -04:00
miloschwartz 98f5e39a7f show ai gateway resource details in launcher 2026-08-11 14:42:07 -04:00
Owen c781be4e70 quiet up some more logs 2026-08-11 14:05:58 -04:00
Owen 48c4b44f72 log chat sessions to database 2026-08-11 12:19:28 -04:00
Owen 68a84181d9 remove excess logging 2026-08-11 12:19:27 -04:00
Owen b60390d805 batch up usage records for insert 2026-08-11 12:19:27 -04:00
miloschwartz 6d45486bb5 add manual virtual api key create ui 2026-08-10 21:44:38 -04:00
Owen 4c1f7f6243 add caching to budget check 2026-08-10 17:39:00 -04:00
miloschwartz 1e3a9fb921 add virtual api key schema and crud endpoints 2026-08-10 17:20:06 -04:00
Owen 8b8e7913dc budget enforcement logic 2026-08-10 17:17:24 -04:00
miloschwartz 93dea0525e basic provider model selector 2026-08-10 12:32:17 -04:00
miloschwartz 187936e5dd use class with helper methods for model catelog 2026-08-10 11:54:57 -04:00
Owen fe5831eb48 pull the models from the upstream server 2026-08-10 09:57:00 -04:00
Owen ca79abc9d4 Update structure 2026-08-10 09:57:00 -04:00
miloschwartz 9eafa067b9 improved provider picker ui on resource 2026-08-07 16:13:51 -04:00
miloschwartz 5e5e172d39 add streaming function to capability 2026-08-07 14:36:15 -04:00
Owen 12056aebc6 Add basic cost calculations for testing 2026-08-07 13:54:32 -04:00
miloschwartz 297cb9c8f2 set provider default capabilities 2026-08-07 12:25:29 -04:00
miloschwartz d699455b38 support changing capabilities on provider integrations 2026-08-07 11:23:07 -04:00
miloschwartz 9e7b4afaec support allow list and blocklist 2026-08-07 10:59:50 -04:00
Owen 22f2990f56 support streaming and closes properly on site targets 2026-08-06 17:15:11 -04:00
miloschwartz c5d68675c9 support patterns in model key 2026-08-06 14:27:08 -04:00
Owen 6cca5e0472 Use fulldoman instead of the alias for the site resources 2026-08-06 13:45:51 -04:00
Owen 75ce7e91d7 Restrict the ai site resource router to exit node subnet 2026-08-06 11:00:34 -04:00
miloschwartz aba27a7bbf support skip tls per request 2026-08-05 17:53:28 -04:00
miloschwartz 3d7e322bf9 add headers to provider 2026-08-05 17:42:58 -04:00
miloschwartz 39e06f2b6d add api capabilities 2026-08-05 16:55:48 -04:00
miloschwartz bcf6b86b84 add no auth and passthrough auth 2026-08-05 15:39:08 -04:00
miloschwartz 2e9bd50172 add provider specific auth modes 2026-08-05 15:17:50 -04:00
Owen 3d062389e9 update oss with gateway traefik routers 2026-08-05 10:14:40 -04:00
Owen 6526d7f225 Remove completions endpoints 2026-08-04 17:09:44 -04:00
Owen aad26b9ae4 move the ai gateway to its own server 2026-08-04 17:08:59 -04:00
Owen 80dcdfe251 change override to work for bother badger and ai gateway 2026-08-04 17:08:59 -04:00
miloschwartz 7759d87835 add basic ui for private inference resource 2026-08-04 16:54:23 -04:00
miloschwartz c085de1e9e Merge branch 'aig' of https://github.com/fosrl/pangolin into aig
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-08-04 15:56:27 -04:00
miloschwartz e38359c74f add crud for adding providers and models to resources 2026-08-04 15:54:24 -04:00
Owen 2cfd7e867b delete the client peers from the exit nodes properly 2026-08-04 15:46:22 -04:00
Owen 9811492a0b send the aliases when connecting for the first time 2026-08-04 14:23:25 -04:00
Owen 0b30cfc341 Working on inference resource management 2026-08-04 11:48:36 -04:00
Owen 1073011a2a first pass of traefik -> basic gateway 2026-08-04 10:15:49 -04:00
miloschwartz 6fa0009ebf add targets and refactor endpoints 2026-08-04 10:15:49 -04:00
Owen 56e59a93f3 Handle the alias create and update special case exit nodes 2026-08-04 10:15:48 -04:00
miloschwartz e5ac6ec7cd add default providers and support overrides 2026-08-04 10:15:48 -04:00
miloschwartz 694fe98131 add crud for providers and models 2026-08-04 10:15:48 -04:00
Owen 32ac8db803 Handle aliases when registering 2026-08-04 10:15:48 -04:00
Owen a790fed297 sync the exit node connection 2026-08-04 10:15:48 -04:00
Owen 33dd10c670 Rename subnet for clarity, pick subnet on client 2026-08-04 10:15:47 -04:00
Owen ba24e1c4f5 Add exit node selection to the clients 2026-08-04 10:15:47 -04:00