Microsoft/bcquality
1
0
Fork 0
mirror of https://github.com/microsoft/BCQuality.git synced 2026-10-05 14:46:55 +01:00
Code Issues Projects Releases Packages Wiki Activity Actions 4

feat(community/agents): add AL agent quality guidance

Browse source 
- add 20 agent knowledge rules with good and bad AL samples
- clarify setup dialog shape, temporary persistence, permissions, profiles, instructions, capability registration, and interface wiring
- add the community-owned AL agents review skill
- make review fixture discovery layer-aware with custom, community, and Microsoft precedence
- document layer-aware evaluation behavior
This commit is contained in:
Stefano Demiliani 2026-08-23 22:10:10 +02:00
parent 841b4e7cab
commit 2d0913568e
63 changed files with 1489 additions and 34 deletions
Download patch file Download diff file Expand all files Collapse all files

6
evaluation/README.md
View file

@ -1,6 +1,6 @@
# AL review evaluation
The evaluation is convention-driven. For every `microsoft/skills/review/al-<domain>-review.md` leaf, the harness finds `microsoft/knowledge/<domain>/`, selects the first article (by filename) with both `.bad.al` and `.good.al` companions, and derives the expected positive and clean control automatically. Adding a conforming leaf requires no scoring-contract edit.
The evaluation is convention-driven. The harness discovers every `<layer>/skills/review/al-<domain>-review.md` leaf across the enabled `microsoft`, `community`, and `custom` layers. Duplicate domains resolve with `custom > community > microsoft` precedence. For each selected leaf, the harness finds paired knowledge across the same layers, applies the same precedence to duplicate article slugs, selects the first article (by filename) with both `.bad.al` and `.good.al` companions, and derives the expected positive and clean control automatically. Adding a conforming leaf requires no scoring-contract edit.
`review-fixtures.json` contains only global thresholds and optional exceptional overrides. An override may select a different article or add context when the generic convention cannot express a scenario. It should remain empty in the normal case.
@ -12,7 +12,7 @@ Model-facing preparation hashes case IDs, neutralizes `Good`/`Bad` object-name t
pwsh ./tools/Test-ReviewFixtures.ps1 -Root .
```
This credential-free check proves every registered leaf maps to a same-named knowledge domain with at least one complete AL sample pair and that all configured overrides are valid.
This credential-free check proves every selected leaf maps to a same-named knowledge domain with at least one complete AL sample pair and that all configured overrides are valid.
## Run a fast-model evaluation
@ -26,7 +26,7 @@ This credential-free check proves every registered leaf maps to a same-named kno
2. For a fast/small model, use one fresh invocation per `request-case-*.json`. Each request embeds the exact leaf instructions, that domain's candidate index rows with authoritative paths, and one opaque case. The model opens only matching articles and copies finding IDs from `candidateArticles[].path`. Save each response with the matching `result-case-*.json` name in the same directory.
`request-<domain>.json` files provide optional two-case leaf batches; save those as `result-<domain>.json`. Directory scoring prefers `result-case-*.json` when present and otherwise falls back to `result-*.json`. `review-request.json` is an optional all-domains stress test for larger models. Neither batch form is the preferred fast-model profile.
`request-<domain>.json` files provide optional two-case leaf batches and identify the selected layer-owned skill path; save those as `result-<domain>.json`. Directory scoring prefers `result-case-*.json` when present and otherwise falls back to `result-*.json`. `review-request.json` is an optional all-domains stress test for larger models. Neither batch form is the preferred fast-model profile.
3. Save only this result shape: