|
|
fa02791fe9
|
Split prompt and profile load errors
|
2026-07-04 23:09:23 +00:00 |
|
|
|
d60ef66f53
|
Implement a library profile API and built-in profile docs
|
2026-07-04 17:23:34 -05:00 |
|
|
|
3ad247039b
|
Add request API key support
|
2026-07-04 16:55:01 +00:00 |
|
|
|
23872dd742
|
Implement runtime parameter completion fixes
ci/woodpecker/tag/release Pipeline was successful
|
2026-07-04 09:00:19 -05:00 |
|
|
|
0512995931
|
Allow JSON-compatible extra params
|
2026-07-04 13:24:27 +00:00 |
|
|
|
049a5feadb
|
Make request execution overrides presence-aware
|
2026-07-04 13:20:39 +00:00 |
|
|
|
63fb8fc132
|
Implement support for OpenRouter sticky routing via a session_id variable
|
2026-07-02 20:08:44 -05:00 |
|
|
|
c95d6fcfec
|
Preserve cache control in rendered prompts
|
2026-07-02 23:03:50 +00:00 |
|
|
|
3ca14d8b6e
|
Add regression test for JSON schema load failure during prepare
|
2026-05-26 13:16:26 +00:00 |
|
|
|
099e9c4a3e
|
Use stable usecase sentinels for HTTP invalid-request mapping
|
2026-05-26 13:14:57 +00:00 |
|
|
|
79901fbb86
|
Refine execution target mapping helpers and coverage across usecase, HTTP, and LLM
|
2026-05-26 13:07:35 +00:00 |
|
|
|
c6c5e3cb69
|
Added support for the service_tier key in profiles
|
2026-05-26 07:44:49 -05:00 |
|
|
|
2091b58066
|
Completed the documentation update and removed the completed roadmap
|
2026-05-26 07:15:36 -05:00 |
|
|
|
f3e8c960af
|
Added support for OpenAI-compatible structured output
|
2026-05-08 07:32:54 -05:00 |
|
|
|
44056d7b8f
|
Add Runner.Prepare and refactor Runner.Run to reuse pre-LLM prepare flow
|
2026-05-06 15:12:58 +00:00 |
|
|
|
bf058a046e
|
Documentation cleanup and bugfixes
|
2026-05-05 20:01:45 -05:00 |
|
|
|
ddb4254124
|
Centralize remaining runtime constants and add precedence/default fallback coverage
|
2026-05-05 11:01:44 -05:00 |
|
|
|
d82bcd0581
|
Migrate repo fixtures to separated prompts/ and execution profiles/ with file-backed prompt templates
|
2026-05-05 10:58:22 -05:00 |
|
|
|
8818796b8b
|
Refactor: separate prompt/profile run resolution with explicit precedence, centralized defaults, and api_key_env validation
|
2026-05-05 10:44:56 -05:00 |
|
|
|
7fffdaede3
|
Refactor: load execution profiles from YAML and split prompt definitions into promptdef repository
|
2026-05-05 10:18:28 -05:00 |
|
|
|
a633c67538
|
Refactor: split prompt definition from execution settings and migrate run contracts to prompt_* + execution_target
|
2026-05-05 10:09:31 -05:00 |
|
|
|
ca4d939fdc
|
Relaxed CLI requirements when defaults are specified in the profile or application defaults
|
2026-05-05 08:41:07 -05:00 |
|
|
|
281202e313
|
Architectural improvements in the http adapter
|
2026-05-05 08:25:32 -05:00 |
|
|
|
c5ce270090
|
Added a configurable timeout knob
|
2026-05-04 22:38:20 -05:00 |
|
|
|
ea05945457
|
Add bounded structured-output repair
|
2026-05-04 21:35:31 -05:00 |
|
|
|
328703df36
|
Add example profiles, schemas, and an end-to-end local fixture
|
2026-05-04 21:23:48 -05:00 |
|
|
|
eac6b69217
|
Add a minimal HTTP API
|
2026-05-04 21:19:29 -05:00 |
|
|
|
ff7e5d3190
|
Implement the core Run use case with a fake LLM path
|
2026-05-04 20:57:08 -05:00 |
|
