-
7124efeb0e
fix(sleep): do not strip regex args in the miner
Christopher Haugen
2026-08-05 09:07:18 +02:00
-
9f434552e9
fix(model): keep a real deployment fallback for copilot; capture timeout stderr
Christopher Haugen
2026-08-05 09:01:56 +02:00
-
f060117219
refactor(sleep): share one char-bound parser between miner and validator
Christopher Haugen
2026-08-05 08:56:20 +02:00
-
bd8e7860ea
fix(model): keep CLI output in the copilot_exec no-response error
Christopher Haugen
2026-08-05 08:52:15 +02:00
-
53d9d9174c
fix(sleep): align mined check args with validate_checks
Christopher Haugen
2026-08-05 08:50:11 +02:00
-
9ec24fc396
fix(model): normalize COPILOT_EXEC_ALLOW_ALL_TOOLS env var on load
Christopher Haugen
2026-08-05 08:31:48 +02:00
-
9b13be0301
test(sleep): compare dispatched store against the normalized config value
Christopher Haugen
2026-08-05 08:29:16 +02:00
-
85a45d1434
fix(sleep): strip refusal markers before bounding the head
Christopher Haugen
2026-08-05 08:28:24 +02:00
-
95d5643db2
fix(model): canonical backend name for fully-local copilot_chat
Christopher Haugen
2026-08-05 08:10:49 +02:00
-
2961a781ab
fix(scripts): treat utf_8 spelling as already-UTF-8 in the console helper
Christopher Haugen
2026-08-05 08:09:10 +02:00
-
453abbe964
fix(sleep): detect markdown-formatted refusals; tighten op-warning test
Christopher Haugen
2026-08-05 08:07:56 +02:00
-
df08665cab
docs(sleep): accurate wording for the leaked-holdout report
Christopher Haugen
2026-08-05 08:06:20 +02:00
-
331d6d700b
fix(sleep): harden harvester further per re-review
Christopher Haugen
2026-08-05 08:02:56 +02:00
-
c052faa73e
fix(sleep): harden Copilot CLI harvester per review follow-ups
Christopher Haugen
2026-08-04 19:00:09 +02:00
-
a44d07b336
fix(sleep): suppress not-validated banner in greedy mode; rename test
Christopher Haugen
2026-08-04 18:57:25 +02:00
-
1f34187580
fix(sleep): guard rubric type and validate mined checks
Christopher Haugen
2026-08-04 18:55:27 +02:00
-
6fde04a412
refactor(scripts): extract UTF-8 console setup into a shared helper
Christopher Haugen
2026-08-04 18:54:04 +02:00
-
cb551d99ae
fix(backend): honor --backend claude when roles are pinned to openai_chat
Christopher Haugen
2026-08-04 18:52:35 +02:00
-
fb93039124
fix(sleep): address review on Copilot CLI harvester
Christopher Haugen
2026-08-04 18:44:17 +02:00
-
b6e726aeea
docs(sleep): correct _split docstring for a lone test-split task
Christopher Haugen
2026-08-04 15:55:01 +02:00
-
56ff307d96
fix(sleep): address review on outcome judges
Christopher Haugen
2026-08-04 15:53:54 +02:00
-
ea40830e62
fix(scripts): address review - best-effort UTF-8 reconfigure, skip already-utf8 streams
Christopher Haugen
2026-08-04 15:51:29 +02:00
-
315deb1dac
feat(model): add the GitHub Copilot CLI as a backend
Christopher Haugen (from Dev Box)
2026-08-04 15:12:12 +02:00
-
adde53850c
feat(sleep): harvest GitHub Copilot CLI sessions
Christopher Haugen (from Dev Box)
2026-08-04 15:13:52 +02:00
-
-
-
6bac6f408b
fix(sleep): never certify a gate verdict measured on the training tasks
Christopher Haugen (from Dev Box)
2026-08-04 15:03:54 +02:00
-
-
-
ae1e722efb
fix(sleep): grade mined tasks on outcome instead of literal strings
Christopher Haugen (from Dev Box)
2026-08-04 15:05:37 +02:00
-
-
-
36d9a9885c
fix(scripts): force UTF-8 on stdout/stderr so runs survive a cp1252 console
Christopher Haugen (from Dev Box)
2026-08-04 14:57:29 +02:00
-
-
-
a971677430
fix(backend): honor --backend when the base config already sets role backends
Christopher Haugen (from Dev Box)
2026-08-04 15:02:44 +02:00
-
-
-
6c74fb895e
test(sleep): make the writes-no-files test able to fail
Dan Baciu
2026-08-03 05:15:04 +04:00
-
5ba2bff1d0
docs(sleep): state the real consolidate_groups return contract
Dan Baciu
2026-08-03 02:57:00 +04:00
-
cbaf6a39f3
feat(sleep): consolidate skill groups independently with failure isolation
Devin AI
2026-07-28 16:36:46 +00:00
-
-
-
7dcc8bf3d5
test(sleep): make the resolver suite portable off POSIX
Dan Baciu
2026-08-03 01:52:45 +04:00
-
a3d3d383a7
fix(sleep): stop a prerelease outranking its stable release
Dan Baciu
2026-08-03 00:50:59 +04:00
-
109bafb44c
feat(sleep): discover the versioned marketplace plugin layout
Dan Baciu
2026-08-03 00:27:32 +04:00
-
1be51ce91f
fix(sleep): address skill-resolver review feedback
Dan Baciu
2026-08-02 22:23:25 +04:00
-
e5d2b241df
feat(sleep): resolve discovered skill names through bounded local roots
Devin AI
2026-07-28 16:34:28 +00:00
-
-
8a4c96a236
Merge pull request #83 from auspic7/feat/pi-source
Yifan Yang
2026-08-03 03:25:33 +08:00
-
-
716197eb18
docs(sleep): list Azure backend in CLI summary
Yif-Yang
2026-08-02 19:24:27 +00:00
-
5b63e1c2be
fix(sleep): harden Pi integration for merge
Yif-Yang
2026-08-02 19:18:50 +00:00
-
6ca6808c82
Merge upstream/main into feat/pi-source
Yif-Yang
2026-08-02 18:43:04 +00:00
-
-
-
-
07d5d42cf1
Merge pull request #196 from Yif-Yang/fix/group-task-copy-normalization
Yifan Yang
2026-08-03 01:58:03 +08:00
-
-
4e1ac00f78
fix(sleep): avoid mutating tasks while grouping
Yif-Yang
2026-08-02 17:57:12 +00:00
-
-
57b9d9d9cf
Merge pull request #184 from bogdanbaciu21/skoc-003-group-tasks-by-skill-hint
Yifan Yang
2026-08-03 01:53:07 +08:00
-
-
48833edb29
feat(sleep): group mined tasks by skill hint with catch-all fallback
Devin AI
2026-07-28 16:31:53 +00:00
-
-
e7014cd18a
Merge pull request #193 from Yif-Yang/fix/copilot-harvest-edge-cases
Yifan Yang
2026-08-02 00:59:43 +08:00
-
-
925d76222a
fix(sleep): harden Copilot transcript harvesting
Yifan Yang
2026-08-01 16:59:08 +00:00
-
-
548af6576b
Merge pull request #191 from asrajawat/feat/skillOpt
Yifan Yang
2026-08-02 00:39:44 +08:00
-
-
7c50b68114
Merge pull request #183 from bogdanbaciu21/skoc-002-propagate-skill-hint-to-taskrecord
Yifan Yang
2026-08-02 00:39:29 +08:00
-
-
da62e7384f
Merge upstream/main into feat/skillOpt
Ashutosh Rajawat
2026-08-01 19:06:08 +05:30
-
-
-
-
4801bcbc0c
fix(sleep): wire copilot source through cli/config/router
Ashutosh Rajawat
2026-08-01 16:18:37 +05:30
-
ee814e6a39
Potential fix for pull request finding
Ashutosh Singh Rajawat
2026-07-31 15:41:09 +05:30
-
5b476ef334
feat(sleep):Support harvest from Vscode copilot sessions #97
Ashutosh Rajawat
2026-07-31 15:14:51 +05:30
-
74c70576b7
fix(sleep): address skill hint review feedback (#183)
Dan Baciu
2026-07-30 10:46:28 +04:00
-
7da46ae693
Merge PR #182: harvest Claude Skill tool invocations
Yifan Yang
2026-07-30 04:57:41 +08:00
-
-
e3437fd7b2
feat(sleep): propagate harvested skill hints into mined tasks
Devin AI
2026-07-28 16:29:50 +00:00
-
-
7b35092a43
feat(sleep): harvest Claude Skill tool invocations
Bogdan Baciu
2026-07-28 16:25:34 +00:00
-
-
8304e6c3ec
Revert "Add ProTeGi vs SkillOpt experiment report"
carpedkm
2026-07-28 09:34:13 +00:00
-
98e767c8da
Add ProTeGi vs SkillOpt experiment report
carpedkm
2026-07-28 09:28:03 +00:00
-
374c832c5a
Merge pull request #181 from Yif-Yang/fix/superpowers-evaluator-hardening
Yifan Yang
2026-07-28 02:50:37 +08:00
-
-
f5c15ecbc2
fix(sleep): harden Superpowers evaluator evidence
Yif-Yang
2026-07-27 18:49:45 +00:00
-
-
59faf00c98
Merge pull request #134 from NovusEdge/feat/superpowers-adapter
Yifan Yang
2026-07-28 02:23:33 +08:00
-
-
c190459b88
Merge pull request #180 from Yif-Yang/docs/blog-main-citation
Yifan Yang
2026-07-28 00:56:27 +08:00
-
-
8b6dbb3a3f
docs(blog): cite the main SkillOpt paper only
Yif-Yang
2026-07-27 16:55:42 +00:00
-
-
bb33bade23
Merge pull request #179 from Yif-Yang/fix/post-merge-sleep-correctness
Yifan Yang
2026-07-27 16:07:08 +08:00
-
-
5c46b2f399
Merge pull request #178 from Yif-Yang/fix/post-merge-security
Yifan Yang
2026-07-27 16:06:51 +08:00
-
-
9acdc0eab6
fix(adapters): correct judge gates, scope superpowers adapter to trusted candidates
NovusEdge
2026-07-27 01:38:51 +03:00
-
74b7f87f4c
fix: close remaining generated-code secret leaks
Yif-Yang
2026-07-26 17:04:05 +00:00
-
-
28efe2ae85
fix(sleep): make validation and model diagnostics truthful
Yif-Yang
2026-07-26 17:06:48 +00:00
-
-
09c6dab5a0
Merge pull request #177 from Yif-Yang/maint/post-merge-hardening
Yifan Yang
2026-07-26 23:19:46 +08:00
-
-
2d658ee234
fix(sleep): reject malformed judges and empty delete anchors
Yif-Yang
2026-07-26 15:17:54 +00:00
-
fa8ce5a041
fix(sleep): track role-specific model and credential settings
Yif-Yang
2026-07-26 15:17:54 +00:00
-
fbf92ca6b5
fix(sleep): preserve context while redacting credentials
Yif-Yang
2026-07-26 15:17:54 +00:00
-
d2b569be30
fix(spreadsheet): close ReAct Python allow-list gaps
Yif-Yang
2026-07-26 15:17:54 +00:00
-
39df792d31
fix(spreadsheet): complete generated-code env isolation
Yif-Yang
2026-07-26 15:17:54 +00:00
-
b9b6d8e019
chore: drop unrelated OSC sync metadata
Yif-Yang
2026-07-26 15:17:54 +00:00
-
-
d6421f903f
Merge pull request #173 from vedmalex/fix/rule-judge-silent-failures
Yifan Yang
2026-07-26 23:18:59 +08:00
-
-
566c0ae85b
Merge pull request #172 from mchillakuru/security/model-change-and-docs
Yifan Yang
2026-07-26 23:18:56 +08:00
-
-
c1ee14a384
Merge pull request #171 from mchillakuru/security/staging-redaction-patterns
Yifan Yang
2026-07-26 23:18:52 +08:00
-
-
b3d6b7dd83
Merge pull request #169 from mchillakuru/security/react-agent-no-shell
Yifan Yang
2026-07-26 23:18:49 +08:00
-
-
a73ab9ee1b
Merge pull request #168 from mchillakuru/security/subprocess-env-isolation
Yifan Yang
2026-07-26 23:18:46 +08:00
-
-
cb59531ebe
Merge pull request #176 from mvanhorn/fix/158-sync-livemath-split-metadata
Yifan Yang
2026-07-26 23:18:26 +08:00
-
-
cd63ada183
fix: sync LiveMath split metadata
Matt Van Horn
2026-07-26 03:00:47 -07:00
-
-
5e3270f611
fix(sleep): address review — surface unmatched edits in the report
vedmalex
2026-07-26 11:20:33 +03:00
-
7348414f7d
fix(sleep): surface silent rule-judge and edit failures
vedmalex
2026-07-26 11:08:36 +03:00
-
ed0d13635b
Move item 202512:46 from val to test split
carpedkm
2026-07-26 07:05:59 +00:00
-
-
-
4a65cc5b2c
security: warn on model swap between nights; correct docs and CLI credential guidance
Murali Chillakuru
2026-07-25 18:46:14 -04:00
-
-
-
609b8d4ac9
security: redact Azure SAS, storage keys, and DB passwords in staging
Murali Chillakuru
2026-07-25 18:43:58 -04:00
-
-
-
085e005d92
security: remove shell=True from ReAct agent bash tool
Murali Chillakuru
2026-07-25 18:41:23 -04:00
-
-
-
4779de76d8
security: scrub subprocess environment for generated code
Murali Chillakuru
2026-07-25 18:40:06 -04:00
-
-
fdeebaf976
Merge pull request #167 from Yif-Yang/docs/media-coverage
Yifan Yang
2026-07-24 16:16:27 +08:00
-
-
0c6ea744ab
docs: add SkillOpt media coverage to News
Yif-Yang
2026-07-24 08:15:35 +00:00
-
-
bd350aa493
Merge pull request #165 from stantheman0128/fix/164-consolidate-test-split-leak
Yifan Yang
2026-07-24 14:04:00 +08:00
-
-
5f8e85c68e
fix(sleep): keep all-test batches out of consolidate train/val
stantheman0128
2026-07-24 00:05:34 +08:00
-
-
640ce85104
fix(adapters): reject symlinked candidate, fail fast on docker without shim python (Copilot round 10)
NovusEdge
2026-07-23 03:34:39 +03:00
-
92fff3e42d
fix(adapters): symlink-safe overlay, is_file candidate check, fail-closed bootstrap (Copilot rounds 8-9)
NovusEdge
2026-07-23 03:29:08 +03:00
-
5e4fb10010
fix(adapters): negation-aware judge, timeout passthrough, docker uid, diagnostics (Copilot rounds 6-7)
NovusEdge
2026-07-23 03:15:16 +03:00
-
e99edbb131
docs(adapters): mark SKILLOPT_SANDBOX experimental; wire SKILLOPT_SHIM_PYTHON
NovusEdge
2026-07-23 03:08:13 +03:00
-
4322656803
fix(adapters): sandbox the verification re-run; claude bin override (Copilot round 5)
NovusEdge
2026-07-23 03:02:59 +03:00
-
0a898b1630
Potential fix for pull request finding
Aliasgar Khimani
2026-07-23 02:59:42 +03:00
-
49356f8e09
fix(adapters): path-safety, minimal PATH, honest marker (Copilot round 4)
NovusEdge
2026-07-23 02:52:39 +03:00