Commit graph

17 commits

Author SHA1 Message Date
Lars Baunwall
778e93cfc1
Mandatory authorization: Always use a bearer token 2025-10-05 14:00:04 +02:00
Lars Baunwall
5127dc0b7f
Refactor code to make it more readable. Fix model family selection that was not working 2025-08-20 19:57:38 +02:00
Devin AI
c8afe9ee17 refactor(extension): extract config/state/logging/models/messages/http modules (SRP)
Co-Authored-By: Lars Baunwall <larslb@thinkability.dk>
2025-08-12 20:40:55 +00:00
Lars Baunwall
f4e6b390e3
Clean-up the extention code 2025-08-12 22:18:05 +02:00
Devin AI
d87fe971b9 Models: dynamic Copilot model listing via LM API; inference supports explicit model selection with 404 model_not_found when unavailable
Co-Authored-By: Lars Baunwall <larslb@thinkability.dk>
2025-08-12 19:34:44 +00:00
Devin AI
f37d4306bc readJson: include data snippet in JSON parse error
Co-Authored-By: Lars Baunwall <larslb@thinkability.dk>
2025-08-12 19:27:47 +00:00
Devin AI
4fd1026e4d nit: name error in catch for JSON.stringify fallback
Co-Authored-By: Lars Baunwall <larslb@thinkability.dk>
2025-08-12 19:25:10 +00:00
Devin AI
a636147dd9 Address review: fix token label ('Set' vs 'None') and remove redundant verbose check in /healthz logging.
Co-Authored-By: Lars Baunwall <larslb@thinkability.dk>
2025-08-12 19:19:58 +00:00
Devin AI
81dd561730 Refactor: migrate Copilot bridge to VS Code Language Model API (vscode.lm). Replace Chat API flow, add robust LM guards and reason codes, update README/manifest, remove Chat API shim.
Co-Authored-By: Lars Baunwall <larslb@thinkability.dk>
2025-08-12 18:46:01 +00:00
Devin AI
71495c6812 Copilot Chat API guard + reason fields: detect missing proposed API, add reason in /healthz and 503 errors, verbose diagnostics, docs; manifest opts into chat proposals
Co-Authored-By: Lars Baunwall <larslb@thinkability.dk>
2025-08-12 18:34:19 +00:00
Devin AI
e61d191ebf Logs/docs: add /healthz verbose log and README ‘Viewing logs’ steps
Co-Authored-By: Lars Baunwall <larslb@thinkability.dk>
2025-08-12 18:25:59 +00:00
Devin AI
862eab2a4a Copilot access auto-reacquire: per-request getAccess(), healthz best-effort under verbose, live status/log updates
Co-Authored-By: Lars Baunwall <larslb@thinkability.dk>
2025-08-12 18:17:07 +00:00
Devin AI
5261239628 Type shim: add minimal VS Code Chat API declarations to compile against proposed API
Co-Authored-By: Lars Baunwall <larslb@thinkability.dk>
2025-08-12 17:46:21 +00:00
Devin AI
e0bca1d50c Cleanup: ensure single throttle guard and request-start instrumentation
Co-Authored-By: Lars Baunwall <larslb@thinkability.dk>
2025-08-12 17:43:49 +00:00
Devin AI
89bb2c0659 Fix try/finally for concurrency, remove duplicate maxConcurrent, add verbose config log
Co-Authored-By: Lars Baunwall <larslb@thinkability.dk>
2025-08-12 17:42:50 +00:00
Devin AI
4bc97d81ac Bridge hardening: 400 payload validation, 429 concurrency cap (configurable), and verbose logging option
Co-Authored-By: Lars Baunwall <larslb@thinkability.dk>
2025-08-12 17:40:07 +00:00
Devin AI
52bd219586 VS Code Copilot Bridge: local OpenAI-compatible HTTP facade to Copilot Chat (SSE + health + models)
Co-Authored-By: Lars Baunwall <larslb@thinkability.dk>
2025-08-12 17:36:15 +00:00