V0.9.2/one prompt eval - #41

Closed

gimlichael wants to merge 36 commits into

mainfrom

v0.9.2/one-prompt-eval

Commits on Sep 7, 2026

💬 clarify eval request workflow and external handoff

aicia-bot
authored and
gimlichael
committed
ea52fe3
🔨 add eval request deterministic helper

aicia-bot
authored and
gimlichael
committed
f683f89
📝 update readme to document eval request capability

aicia-bot
authored and
gimlichael
committed
4a22669
📝 document v0.9.2 release notes for one-shot eval handoff
gimlichael
committed
0bc863e
🐛 guard started eval packages from -Force deletion
aicia-bot
committed
f12d15d
♻️ clarify eval orchestrator capability and guard model policy
aicia-bot
committed
512b503

Commits on Sep 8, 2026

✅ implement fail-closed copilot authentication validation
aicia-bot
committed
eee4240
📝 document copilot authentication incompatibility handling
aicia-bot
committed
3b1703d
💬 document eval infrastructure and project conventions
aicia-bot
committed
7faba4a
👷 enhance eval infrastructure with grading and isolation
aicia-bot
committed
ca663cf
✅ add evaluation configuration for dotnet-change-impact skill
aicia-bot
committed
611b449
✅ add test for eval git workspace scenarios
aicia-bot
committed
944dfac
♻️ clarify git-visual-commits scope and extract grouping guidance
aicia-bot
committed
05e132b
🔨 update skill validation to reference grouping examples
aicia-bot
committed
66d2fdc
💬 document git-visual-commits scope and inventory features
aicia-bot
committed
469b4a3
🐛 fix eval runner cross-platform path and environment support
aicia-bot
committed
ab54ffa
🐛 fix explicit gh_config_dir auth selection

gimlichael
and
Copilot
committed
9cd1c5b
💬 clarify eval git fixture isolation policy

gimlichael
and
Copilot
committed
b9bccbc

Commits on Sep 9, 2026

♻️ improve dotnet-change-impact compatibility classifications
aicia-bot
committed
15c0e02
🔧 enhance evaluation framework with analyzer/grader separation
aicia-bot
committed
5dea7f0
✅ update evaluation test fixtures and validation
aicia-bot
committed
87ba9ba
🏗️ enforce phase 2 analyzer provenance

gimlichael
and
Copilot
committed
f90bc06
📝 document phase 2 analyzer workflow

gimlichael
and
Copilot
committed
e1c0c27
✅ cover analyzer catalog injection

gimlichael
and
Copilot
committed
520cb67
✅ harden analyzer model validation tests

gimlichael
and
Copilot
committed
06e5adb
💚 allow phase 2 validation runtime

gimlichael
and
Copilot
committed
7268bb4
🔧 enhance eval runner infrastructure for transcripts and validation
aicia-bot
committed
359cdd4
✅ update test and validator infrastructure for eval execution
aicia-bot
committed
0cbf72e
🐛 fix eval handoff and transcript evidence checks

aicia-bot
and
Copilot
committed
8b9cd29

Commits on Sep 10, 2026

🐛 bind prepared handoff prompt bytes to the manifest

gimlichael
and
Copilot
committed
7b2c4c8
💬 clarify eval handoff decision workflow and state lifecycle
aicia-bot
committed
aeb7e9d
♻️ formalize eval handoff decision states and validation
aicia-bot
committed
51c5e84
🔧 enhance skill validation templates
aicia-bot
committed
deeb3a9
🏗️ enhance eval-runner conformance and isolation validation
aicia-bot
committed
d1fb99f
✅ add eval-runner conformance and isolation tests
aicia-bot
committed
1ab3e56
♻️ extract path utilities and add containment tests
aicia-bot
committed
df9f405