Repository navigation
Fixes AB#3789817 [ATS Perf] [android-complete] Update PR review skills to prevent performance regressions - #475
Open
Siddhi (siddhijain) wants to merge 2 commits into
Conversation
Add explicit attribution metadata and per-PR outcome reporting, including confirmed zero-finding reviews and regression fixtures. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Copilot-Session: 61a54db1-5bab-4bf4-9f9a-9e478fa4a1dc
|
❌ Invalid work item number: AB#3789817 ##. Work item number must be a valid integer. Click here to learn more. |
Siddhi (siddhijain)
changed the title
|
✅ Work item link check complete. Description contains link AB#3789817 to an Azure Boards work item. |
There was a problem hiding this comment.
🟡 Changes recommended
Quoted Markdown continuation lines can be mistaken for attribution markers, producing false screening evidence and inflated counts.
1 open issue
What changed in this PR
Adds memory-storage/v1 attribution tracking to the review-analysis tooling; Common and Broker guidance changes remain separate.
Changes:
- Parses markers and preserves rule, review, and guidance metadata.
- Generates deduplicated attribution counts, including confirmed zero-finding reviews.
- Adds summary pagination, empty-dataset handling, and mocked regression fixtures.
| File | Description |
|---|---|
.github/skills/copilot-review-analyst/tests/rule-attribution.tests.ps1 |
Adds parsing, classification, and collector fixtures. |
.github/skills/copilot-review-analyst/SKILL.md |
Documents attribution reporting and limitations. |
.github/skills/copilot-review-analyst/scripts/rule-attribution.ps1 |
Adds metadata-marker parsing. |
.github/skills/copilot-review-analyst/scripts/final-classification.ps1 |
Exports attribution counts and handles empty results. |
.github/skills/copilot-review-analyst/scripts/analyze.ps1 |
Collects metadata and paginates review summaries. |
Review effort: Balanced
Track quoted paragraphs and code fences so quoted examples cannot supply rule, screen, or guidance metadata. Cover all three markers and valid top-level markers after quoted content. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Copilot-Session: 61a54db1-5bab-4bf4-9f9a-9e478fa4a1dc
Siddhi (siddhijain)
added a commit
to AzureAD/microsoft-authentication-library-common-for-android
that referenced
this pull request
Oct 22, 2026
…nt performance regressions (#3321) Fixes [AB#3797297](https://identitydivision.visualstudio.com/fac9d424-53d2-45c0-91b5-ef6ba7a6bf26/_workitems/edit/3797297) ## Summary Update Common's actual Copilot code-review instructions and repository-local review skill to check memory and storage performance across every production-code PR, including new features outside authentication and cache code. - Review whole-data reads, repeated decryption/parsing, allocation and copy amplification, collection complexity, cache reuse, retention/invalidation, and expensive work on request/UI paths or under locks. - Trace affected callers and callees, including unchanged helpers newly invoked or called more often, without auditing unrelated legacy code. - Use the ATS improvements as concrete examples: reuse in-memory account/credential records and avoid full token-cache decryption during routine token saves. - Require actionable evidence and preserve encryption, persistence, authorization boundaries, synchronization, freshness, ordering, and defensive-copy guarantees. - Tag matching findings with `Review rule: memory-storage/v1` and record screen completion/guidance revision in review summaries, including zero-finding reviews. ## Attribution The markers integrate with the measurement tooling in [AzureAD/android-complete#475](AzureAD/android-complete#475), which reports rule-associated comments and outcomes. This PR changes the review guidance itself, not the analysis tooling or authentication runtime. The performance screening is useful independently of the reporting rollout. ## Scope and validation Documentation-only changes to three files: - `.github/copilot-instructions.md` - `.github/skills/code-review/SKILL.md` - `.github/skills/code-review/references/performance-resources-dependencies.md` No production code, dependencies, or cache behavior changes. Existing false-positive suppression and comment-quality policies remain in place. No runtime build or tests are required for these instruction changes. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Copilot-Session: 61a54db1-5bab-4bf4-9f9a-9e478fa4a1dc
This branch has not been deployed
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Fixes AB#3789817
Summary
Add attribution tracking to the Copilot review-analysis tooling for the
memory-storage/v1review rule. This is the first PR in the rollout; the Common and Broker review-guidance changes will be reviewed separately.rule_attribution.jsonwith per-repository/PR comment counts, outcome buckets, screen confirmation, and guidance revisions.Counts are deduplicated by repository and comment ID. They measure rule-associated comments, not unique defects or proof that the guidance caused an incremental discovery. Existing outcome-classification and force-push verification requirements remain unchanged.
Validation
Passed locally. The fixtures cover marker parsing, quoted/fenced examples, all outcome buckets, duplicate comment IDs, cross-repository PR numbers, zero-finding and legacy datasets, review-summary pagination, and metadata beyond the summary excerpt. GitHub transport is mocked; the tests do not require network access or credentials.
Scope
Only the five review-analysis guidance, script, and test files are included. Common/Broker module changes and unrelated workspace configuration are excluded.