github/microsoft-SkillOpt
1
0
Fork 0
mirror of https://github.com/microsoft/SkillOpt.git synced 2026-07-03 14:02:58 +08:00
Code Issues Packages Projects Releases Wiki Activity

Commit Graph

  • e4ea6a6771 chore(release): v0.2.0 main v0.2.0 CharlesYang030 2026-07-02 22:11:10 +08:00
  • 5487e2c426 fix(skillopt-sleep): redact secrets before persisting cycle diagnostics Yif Yang 2026-06-30 19:47:36 +00:00
  • b9142bad24 fix(skillopt-sleep): surface codex auth/model/version failures instead of silently scoring 0 (#92) Yifan Yang 2026-07-01 03:20:08 +08:00
  • 95a9e959fe test(sleep): add verifier-discipline stress test for the validation gate (#87) Yifan Yang 2026-07-01 02:40:24 +08:00
  • 680dd28f5a fix(tests): move TestVerifierDiscipline above main block Tanmay9223 2026-06-30 13:05:01 +05:30
  • fccc21f3f6 test(sleep): add verifier-discipline stress test (closes #67) Tanmay9223 2026-06-24 23:54:48 +05:30
  • 6849e609a3 feat(eval): add missing minimax backend configuration Yifan Yang 2026-06-23 20:31:39 +08:00
  • 9fa0716c72 fix(skillopt-sleep): also surface codex failures on the tool-call rollout path Daniel Martinez 2026-06-27 23:56:11 -05:00
  • 9fcf5868c3 fix(skillopt-sleep): surface codex auth/model/version failures instead of silently scoring 0 Daniel Martinez 2026-06-27 22:23:19 -05:00
  • 9969a8f393 Add Devin plugin (plugins/devin): MCP server + ATIF-v1.7 harvest (#88) Yifan Yang 2026-06-26 11:04:23 +08:00
  • 26e5338def Update citation from @misc to @article format Yif Yang 2026-06-26 02:54:46 +00:00
  • 1a70e4c9cd devin harvest: space turns >=5s so single-turn sessions aren't dropped khashayar 2026-06-25 22:03:15 +02:00
  • 9799c41461 devin plugin: full schema/tool parity with plugins/copilot khashayar 2026-06-25 21:56:42 +02:00
  • e51eb7c4be devin plugin: expand ~ in CLAUDE_HOME from env + add tests & ATIF fixture khashayar 2026-06-25 21:49:21 +02:00
  • 99ccb93945 fix(eval-only): configure qwen_chat/minimax backends so local LLM endpoints work (#85) Yifan Yang 2026-06-26 02:55:18 +08:00
  • 9de9220214 docs(sleep): add cross-model scaling results (nano +11.9) and hyperparam ablation (#89) Yifan Yang 2026-06-26 01:40:58 +08:00
  • bec23ed020 Add Devin plugin (plugins/devin): MCP server + ATIF-v1.7 harvest khashayar 2026-06-25 10:42:52 +02:00
  • 8559308361 fix(eval-only): call configure_qwen_chat so itslocal LLM endpoints can be used Gergely Imreh 2026-06-24 15:00:56 +08:00
  • 2d7e37a395 fix(json_utils): reject prose pseudo-JSON in single quotes/backticks (#82) Yifan Yang 2026-06-23 20:31:39 +08:00
  • baad64a3b9 docs(readme): remove Acknowledgements section (#81) Yifan Yang 2026-06-23 19:13:16 +08:00
  • c2e47c50fb docs(readme): acknowledge community contributor @samuelgoofus-boop (#80) Yifan Yang 2026-06-23 19:03:30 +08:00
  • 14c045f04f Windows robustness for claude/codex backends (+ hardened JSON fallback) (#79) Yifan Yang 2026-06-23 19:00:23 +08:00
  • 2841f82428 Fix ALFWorld gamefile paths relative to ALFWORLD_DATA carpedkm 2026-06-23 10:32:38 +00:00
  • 64c6dda105 Merge pull request #78 from Yif-Yang/main Yifan Yang 2026-06-23 16:52:42 +08:00
  • c98eac18c7 docs(readme): add Trendshift daily/weekly badges (#1) Yifan Yang 2026-06-23 16:50:47 +08:00
  • fc1f827f07 Merge pull request #74 from Yif-Yang/fix/python-path-and-lookback Yifan Yang 2026-06-20 22:26:43 +08:00
  • 01b3e01804 fix: use None default for --lookback-hours to distinguish omitted vs 0 carpedkm 2026-06-20 14:23:17 +00:00
  • 01075c90d3 fix: address codex round 2 — revert harvest break + allow lookback 0 carpedkm 2026-06-20 14:21:18 +00:00
  • 6cc1cd2e95 fix: address codex review — use clock for cutoff + early-exit harvest carpedkm 2026-06-20 14:11:58 +00:00
  • 889238b234 fix: add SKILLOPT_SLEEP_PYTHON override + lookback_hours first-run fallback carpedkm 2026-06-20 14:07:50 +00:00
  • b5a1c2b317 Merge pull request #73 from Yif-Yang/fix/bare-subscription-auth Yifan Yang 2026-06-20 21:46:09 +08:00
  • 552ddefd74 fix: narrow CLI error markers to avoid false positives carpedkm 2026-06-20 13:32:43 +00:00
  • bfa53bc46d fix(sleep): make --bare conditional on ANTHROPIC_API_KEY (#68) carpedkm 2026-06-20 13:28:34 +00:00
  • 24b5a25ba8 Merge pull request #72 from Yif-Yang/feat/plugin-feature-sync Yifan Yang 2026-06-20 20:42:24 +08:00
  • 0d648b2580 fix: address codex+gpt-5.5 review findings carpedkm 2026-06-20 12:40:34 +00:00
  • 7d36b1d592 fix: address review findings in plugin sync PR carpedkm 2026-06-20 12:04:07 +00:00
  • 0be780052a feat: sync all 4 runtime plugins with full engine surface + fix #52 #58 #62 carpedkm 2026-06-20 11:31:09 +00:00
  • 0b5b9a4296 Merge pull request #60 from Kirchberg/codex/reviewed-task-files-cwd carpedkm 2026-06-20 08:59:02 +00:00
  • 05cdc26beb Add reviewed task-file flow for Codex sleep runs Kirill Kostarev 2026-06-15 14:45:46 +03:00
  • 382811ddcc Merge pull request #50 from Dongbumlee/Dongbumlee/copilot-sleep-backend Yifan Yang 2026-06-20 16:57:53 +08:00
  • d367ae1eea docs(plugins): list copilot in the cross-tool backend overview DB Lee 2026-06-17 17:38:10 -07:00
  • 2c0980bda3 docs(copilot): correct backend hint in research MCP plugin (openai -> azure_openai) DB Lee 2026-06-12 09:19:29 -07:00
  • 5799695951 feat(copilot): implement attempt_with_tools with cross-platform tool shims DB Lee 2026-06-12 09:05:13 -07:00
  • 013a7cd83a test: add unit tests for CopilotCliBackend (parsing + alias + isolated home) DB Lee 2026-06-12 08:32:46 -07:00
  • 21f93c16c7 Add GitHub Copilot backend to SkillOpt-Sleep DB Lee 2026-06-12 08:21:57 -07:00
  • 5dc894715f Add SkillOpt research-engine MCP server plugin for Copilot DB Lee 2026-06-12 08:21:47 -07:00
  • 6940e46f4e Merge pull request #65 from summerview1997/codex/searchqa-materialize-splits Yifan Yang 2026-06-17 23:50:38 +08:00
  • 0e962219f5 Merge pull request #64 from summerview1997/codex/searchqa-rollout-failfast Yifan Yang 2026-06-17 23:49:55 +08:00
  • fc42e6bf72 Merge pull request #63 from summerview1997/codex/webui-env-backend-preflight Yifan Yang 2026-06-17 23:49:50 +08:00
  • c755792049 Add SearchQA materialization tests summerview1997 2026-06-16 09:27:09 +08:00
  • e591a28242 Add SearchQA split materialization helper summerview1997 2026-06-16 09:26:56 +08:00
  • c04467a428 Add SearchQA materialization dependency extra summerview1997 2026-06-16 09:26:46 +08:00
  • d5ae8c8e66 Document SearchQA split materialization summerview1997 2026-06-16 09:26:35 +08:00
  • 923becb00f Add SearchQA rollout fail-fast tests summerview1997 2026-06-16 09:21:08 +08:00
  • da799620ba Fail fast on systemic SearchQA rollout failures summerview1997 2026-06-16 09:20:57 +08:00
  • 30cc8a3ed3 Add WebUI env preflight tests summerview1997 2026-06-16 09:04:30 +08:00
  • d05851bd7f Add WebUI env loading and backend preflight summerview1997 2026-06-16 09:04:19 +08:00
  • 46b3207b96 docs(sleep): trim RESULTS to the headline results (remove the full grid) Yifan Yang 2026-06-15 17:08:51 +00:00
  • d43e8dba1a docs(sleep): expand the grid into per-benchmark night-by-night tables Yifan Yang 2026-06-15 16:54:01 +00:00
  • d02098ffc4 docs(sleep): add full Results & Analysis (RESULTS.md); link from README Yifan Yang 2026-06-15 16:49:13 +00:00
  • ea4ff459d7 docs(sleep): make the results section rigorous (named benchmarks, baseline→after) Yifan Yang 2026-06-15 16:42:43 +00:00
  • de3be75bac docs(sleep): add a SkillOpt-Sleep module readme + News mention Yifan Yang 2026-06-15 16:31:15 +00:00
  • b701d9b6d9 docs: move SkillOpt-Sleep into the guide; clean docs/sleep; fix guide link Yifan Yang 2026-06-15 16:20:50 +00:00
  • 722ce646d4 feat(sleep): experience replay + dream rollouts in the cycle (opt-in) Yifan Yang 2026-06-15 15:58:27 +00:00
  • 576f2f8bad Merge pull request #59 from Elzlxx/feat/openclaw-skillopt-sleep Yifan Yang 2026-06-15 18:26:12 +08:00
  • 00d07bc59a Merge pull request #48 from Kirchberg/codex/codex-desktop-harvest carpedkm 2026-06-15 10:23:18 +00:00
  • 31715a8b43 Add Codex Desktop transcript harvesting Kirill Kostarev 2026-06-12 16:37:23 +03:00
  • e8c3e10b30 Merge pull request #49 from Kirchberg/codex/codex-skill-first-upstream carpedkm 2026-06-15 10:21:43 +00:00
  • d31e9d9407 Back up legacy Codex prompt during install Kirill Kostarev 2026-06-12 16:58:26 +03:00
  • 1953484822 Make Codex integration skill-first Kirill Kostarev 2026-06-12 16:51:54 +03:00
  • 1b2652c6f8 Merge pull request #44 from imshunsuke/refactor/reflect-default-base carpedkm 2026-06-15 09:06:38 +00:00
  • 98d0430bee refactor: make EnvAdapter.reflect a shared default (fixes dropped reflect kwargs) refactor/reflect-default-base Shunsuke 2026-06-09 18:51:11 +08:00
  • eef4805b25 Merge pull request #43 from imshunsuke/docs/fix-benchmark-loader-naming Yifan Yang 2026-06-15 17:00:45 +08:00
  • 86bad36ffe feat(sleep): SkillOpt-Sleep plugin update (preview) — engine robustness + scheduling Yifan Yang 2026-06-14 16:12:00 +00:00
  • 553446575a feat(plugins): add OpenClaw shell for SkillOpt-Sleep elzlxx 2026-06-14 23:27:54 +08:00
  • c1ac570d94 docs(guideline): make SearchQA the first demo — copy-paste materialization snippet + train command feat/skill-aware-reflection Cuzyoung 2026-06-10 13:48:43 +00:00
  • d8023a47c9 docs(guideline): novice-first restructure — Quick Start before data, honest first-demo path, own-data narrative Cuzyoung 2026-06-10 13:42:50 +00:00
  • b0b62fcb86 docs(readme): slim README — move install/quick-start/data/config details to the guideline page Cuzyoung 2026-06-10 13:27:36 +00:00
  • 3308c4c5dc docs(guideline): add PyPI install option and skill-aware reflection config rows Cuzyoung 2026-06-10 13:27:12 +00:00
  • 0d5b331cd5 Merge branch 'docs/guideline' into feat/skill-aware-reflection Cuzyoung 2026-06-10 13:27:12 +00:00
  • 1c6a0e75c8 docs(guide): document skill-aware reflection options in the configuration guide Cuzyoung 2026-06-10 13:19:27 +00:00
  • 88989d120d chore: ignore local experiment launcher scripts (machine-specific endpoints/identities) Cuzyoung 2026-06-10 13:10:55 +00:00
  • 44043d4ae5 docs(trainer): drop the stale skill-aware comments (claimed best_skill carries no appendix; it does) Cuzyoung 2026-06-10 12:06:05 +00:00
  • 7dcd612361 fix(trainer): flush appendix notes on skip branches — lapse-only steps no longer drop them Cuzyoung 2026-06-10 11:31:03 +00:00
  • 0dc84162dc feat(optimizer): skill-aware reflection (EmbodiSkill S_app), config-controlled and env-independent Cuzyoung 2026-06-10 11:28:29 +00:00
  • ffe581098b feat(trainer): final-skill val + best promotion; keep best unpolluted by slow_update Cuzyoung 2026-06-02 05:55:31 +00:00
  • 372fd56c1e fix(spreadsheetbench)+optimizer: fix verify-feedback bloat, drop optimizer-side truncation, soft-disable gate Cuzyoung 2026-06-01 11:23:08 +00:00
  • 54e4b3eafb docs: align benchmark guide and template with dataloader.py naming Shunsuke 2026-06-09 12:20:01 +08:00
  • f64a41397c docs(sleep): add PR draft (title + body) for the upstream PR Yifan Yang 2026-06-08 14:31:52 +00:00
  • 5cd22bb71b docs: add PUBLISHING.md — how users install the three plugins Yifan Yang 2026-06-08 14:31:52 +00:00
  • d6c4ca3f6e docs(sleep): load-test all 3 plugin shells on a fresh (non-gbrain) example Yifan Yang 2026-06-08 14:31:52 +00:00
  • dae974a5e3 chore(sleep): English-only across the engine, plugins, and docs Yifan Yang 2026-06-08 14:31:52 +00:00
  • f9db99853b feat(plugins): ship SkillOpt-Sleep for Claude Code, Codex, and Copilot Yifan Yang 2026-06-08 14:31:52 +00:00
  • b02ffc2c99 refactor(sleep): decouple engine to top-level skillopt_sleep/ (zero research dep) Yifan Yang 2026-06-08 14:31:52 +00:00
  • e2de84d36f docs(sleep): real Claude<->Codex cross-validation of the new features Yifan Yang 2026-06-08 14:31:51 +00:00
  • 9379e494bf docs(sleep): document the controllable dreaming architecture Yifan Yang 2026-06-08 14:31:51 +00:00
  • a29201adc4 feat(sleep): multi-objective reward (accuracy/tokens/latency) + user preferences Yifan Yang 2026-06-08 14:31:51 +00:00
  • 77ac33e8bf feat(sleep): multi-rollout contrastive reflection + token/time budget Yifan Yang 2026-06-08 14:31:51 +00:00
  • c179a24c45 feat(sleep): slow-update long-term memory field (runs even with gate off) Yifan Yang 2026-06-08 14:31:51 +00:00
  • 6f1351edb9 feat(sleep): 3-way train/val/test split + gate_mode on|off Yifan Yang 2026-06-08 14:31:51 +00:00