onCodeDesign · florinc · Mar 19, 2026 · Mar 19, 2026 · Mar 19, 2026 · Mar 17, 2026
diff --git a/.copilot-cli/1-overdue-orders.coder.cli.md b/.copilot-cli/1-overdue-orders.coder.cli.md
diff --git a/.github/agents/architect.agent.md b/.github/agents/architect.agent.md
@@ -1,7 +1,6 @@
 ---
 description: 'Architecture design agent for plugin-based modular systems following Clean Architecture principles with strict boundary enforcement'
 tools: ['execute/getTerminalOutput', 'execute/runTask', 'execute/getTaskOutput', 'execute/createAndRunTask', 'execute/runInTerminal', 'read/getNotebookSummary', 'read/readFile', 'edit/createDirectory', 'edit/createFile', 'edit/editFiles', 'edit/editNotebook', 'github/issue_read']
-model: Claude Sonnet 4.5 (copilot)
 handoffs:
   - label: Review Architecture Design
     agent: architect

diff --git a/.github/agents/code-reviewer.agent.md b/.github/agents/code-reviewer.agent.md
@@ -1,7 +1,6 @@
 ---
 description: 'Code review agent that verifies implementations match design documents, enforce architectural constraints, and meet quality standards'
 tools: ['execute/getTerminalOutput', 'execute/runTask', 'execute/getTaskOutput', 'execute/createAndRunTask', 'execute/runInTerminal', 'read/getNotebookSummary', 'read/readFile', 'edit/createDirectory', 'edit/createFile', 'edit/editFiles', 'edit/editNotebook', 'github/issue_read']
-model: Claude Sonnet 4.6 (copilot)
 required_skills:
   - path: '.github/skills/code-review-report/SKILL.md'
     when: 'always'
@@ -127,6 +126,7 @@ Enforce rules from `.github/copilot-instructions.md`:
 
 ### 2. Read Implementation
 - Read every file in the provided file list
+- Map each file to the commit_id from the provided commits list for context 
 - For each file, understand its role (contract, service, entity, interceptor, test, etc.)
 - If build/test status failed, note failures upfront
 
@@ -173,9 +173,16 @@ Output a summary as following:
 
 ### Review Report
 
-Output a structured review report in markdown format, and save it at `docs/code-reviews/{issueId}-code-reviewer_{timestamp}.md`.
+Output a structured review report in the requested format. Save it at `docs/code-reviews/{issueId}-code-review_{timestamp}.{extension}`.
+
+Use below table to determine the skill you will use to generate the report:
+
+| Format | Skill | Extension |
+|--------|-------|-----------|
+| markdown | `code-review-md-report` | md |
+| json | `code-review-json-report` | json |
+| yaml | `code-review-yaml-report` | yaml |
 
-Use the `code-review-report` skill to generate the report.
 In case of not being able to use the skill, report a error and produce a simple markdown report.
 
 ## Error Recovery

diff --git a/.github/agents/coder.agent.md b/.github/agents/coder.agent.md
@@ -1,7 +1,6 @@
 ---
 description: 'Implementation agent that converts detailed design specifications into working C# code following Clean Architecture principles and strict dependency rules'
 tools: ['execute/getTerminalOutput', 'execute/runTask', 'execute/getTaskOutput', 'execute/createAndRunTask', 'execute/runInTerminal', 'read/getNotebookSummary', 'read/readFile', 'edit/createDirectory', 'edit/createFile', 'edit/editFiles', 'edit/editNotebook', 'github/issue_read']
-model: Claude Sonnet 4.5 (copilot)
 required_skills:
   - path: '.github/skills/unit-testing/SKILL.md'
     when: 'mode == "unit-tests"'

diff --git a/.github/agents/detailed-designer.agent.md b/.github/agents/detailed-designer.agent.md
@@ -1,7 +1,6 @@
 ---
 description: 'Detailed design agent that converts approved high-level architecture into implementable specifications with explicit contracts, error handling, and test strategies'
 tools: ['execute/getTerminalOutput', 'execute/runTask', 'execute/getTaskOutput', 'execute/createAndRunTask', 'execute/runInTerminal', 'read/getNotebookSummary', 'read/readFile', 'edit/createDirectory', 'edit/createFile', 'edit/editFiles', 'edit/editNotebook', 'github/issue_read']
-model: Claude Sonnet 4.5 (copilot)
 handoffs:
   - label: Review Detailed Design
     agent: detailed-designer

diff --git a/.github/copilot-instructions.md b/.github/copilot-instructions.md
@@ -212,7 +212,7 @@ Domain-specific guidance documents located in `.github/skills/`. When a task mat
 |-------|---------|
 | **add-module-plugin** | Step-by-step guide for adding new modules: folder structure, plugin registration, `IModule` initialization, build-order dependencies in `.sln` |
 | **unit-testing** | Test patterns using xUnit, NSubstitute, FluentAssertions: AAA structure, fake naming (`Stub`/`Mock`), collection assertions, `GetTarget` helpers |
-| **code-review-report** | Structured code review report template for PR-importable Markdown output (GitHub / Azure DevOps compatible) |
+| **code-review-md-report** | Structured code review report template for PR-importable Markdown output (GitHub / Azure DevOps compatible) |
 
 Usage pattern in this file: `> For [topic], use the **skill-name** skill.`
 

diff --git a/.github/skills/code-review-report/SKILL.md → ...hub/skills/code-review-md-report/SKILL.md b/.github/skills/code-review-report/SKILL.md → ...hub/skills/code-review-md-report/SKILL.md
@@ -1,47 +1,40 @@
 ---
-name: code-review-report
-description: "Generate structured code review reports in Markdown format suitable for import into GitHub PRs or Azure DevOps pull requests. Produces per-file inline remarks and a summary that renders natively in PR comment threads."
-version: 2.0.0
-output_format: Markdown + JSON
-platforms: GitHub, Azure DevOps
+name: code-review-md-report
+description: "Generate structured code review reports in Markdown format. Produces per-file inline remarks and a summary"
+version: 1.0.0
+output_format: Markdown
 ---
 
-# Code Review Report Skill
+# Code Review Markdown Report Skill
 
 ## Purpose
 
-Generates a structured, portable code review report consisting of **two artifacts**:
+Generates a structured, portable code review report consisting of **one artifact**:
 
 1. **Markdown report** — human-readable, renders correctly as a GitHub or Azure DevOps PR comment; can be pasted directly or posted via `gh pr comment`
-2. **JSON payload** — machine-ready structured data for programmatic import via `gh api` (GitHub) or `curl` (Azure DevOps REST API)
 
-Both artifacts are saved to `docs/code-reviews/` and referenced in the chat summary.
+The artifact is saved to `docs/code-reviews/` and referenced in the chat summary.
 
 ## When to Use
 
 - After completing a code review with the `code-reviewer` agent
-- When the review output must be shared in a pull request on GitHub or Azure DevOps
+- When the review output is requested as a markdown report (e.g., "Generate a markdown code review report ...")
 - When a persistent review artifact is needed for traceability
 
-## Output Files
+## Output File
 
-| File | Path | Purpose |
-|------|------|---------|
-| Markdown report | `docs/code-reviews/{issueId}-review_{timestamp}.md` | Human-readable, paste into PR comment |
-| JSON payload | `docs/code-reviews/{issueId}-review_{timestamp}.json` | Programmatic PR import (GitHub + Azure DevOps) |
+Markdown report: `docs/code-reviews/{issueId}-code-review_{timestamp}.md` - Human-readable, paste into PR comment
 
 **Timestamp format:** `yyyyMMdd-HHmm` (e.g., `20260225-1430`)
 
-> Generate the JSON payload **only** when at least one remark has a resolvable file path. If all remarks are general (no file/line), produce only the Markdown report.
-
-## Artifact 1 — Markdown Report Template
+## Markdown Report Template
 
 Use the following template **exactly** as the structure for the report file. Replace all `{placeholders}` with actual values. Omit sections that have no content (e.g., if no blockers, omit the blockers subsection) but always keep the top-level sections.
 
 ---
 
 ```markdown
-# Code Review Report
+# Code Review Markdown Report
 
 | Field | Value |
 |-------|-------|
@@ -87,6 +80,7 @@ Use the following template **exactly** as the structure for the report file. Rep
 #### R{number}: {short title}
 
 - **File:** `{relative/path/to/File.cs}`
+- **Commit:** `{commit-id}`
 - **Line(s):** {line or range}
 - **Dimension:** {Design Conformance | Architecture Compliance | Code Quality | Test Quality | Scope & Hygiene}
 - **Description:** {What is wrong and why it matters}
@@ -101,6 +95,7 @@ Use the following template **exactly** as the structure for the report file. Rep
 #### R{number}: {short title}
 
 - **File:** `{relative/path/to/File.cs}`
+- **Commit:** `{commit-id}`
 - **Line(s):** {line or range}
 - **Dimension:** {dimension}
 - **Description:** {description}
@@ -115,6 +110,7 @@ Use the following template **exactly** as the structure for the report file. Rep
 #### R{number}: {short title}
 
 - **File:** `{relative/path/to/File.cs}`
+- **Commit:** `{commit-id}`
 - **Line(s):** {line or range}
 - **Dimension:** {dimension}
 - **Description:** {description}
@@ -162,115 +158,6 @@ Use the following template **exactly** as the structure for the report file. Rep
 
 ---
 
-## Artifact 2 — JSON Payload Template
-
-The JSON payload serves a dual purpose: it is the body for the **GitHub PR Review API** and contains all data needed to reconstruct **Azure DevOps PR threads**. The agent fills this structure from the collected remarks.
-
-```json
-{
-  "_meta": {
-    "issueId": "{issueId}",
-    "issueTitle": "{issueTitle}",
-    "date": "{date}",
-    "timestamp": "{timestamp}",
-    "markdownReport": "docs/code-reviews/{issueId}-review_{timestamp}.md"
-  },
-  "verdict": "{APPROVE|REQUEST_CHANGES|COMMENT}",
-  "summary": "{2-4 sentence overview}",
-  "body": "{full markdown summary block — same as the Markdown report's Summary + Metrics + Verdict sections}",
-  "event": "{APPROVE|REQUEST_CHANGES|COMMENT}",
-  "metrics": {
-    "blockers": 0,
-    "warnings": 0,
-    "suggestions": 0,
-    "notes": 0
-  },
-  "remarks": [
-    {
-      "id": "R1",
-      "severity": "blocker",
-      "dimension": "Architecture Compliance",
-      "file": "Modules/Sales/Sales.Services/OrderingService.cs",
-      "line": 42,
-      "lineEnd": 42,
-      "title": "{short title}",
-      "description": "{full description}",
-      "suggestedFix": "{direction — no full code}"
-    }
-  ]
-}
-```
-
-### JSON Field Rules
-
-| Field | Values | Notes |
-|-------|--------|-------|
-| `verdict` / `event` | `APPROVE`, `REQUEST_CHANGES`, `COMMENT` | `APPROVE WITH SUGGESTIONS` → use `COMMENT` |
-| `severity` | `blocker`, `warning`, `suggestion`, `note` | Lowercase |
-| `file` | Repo-root relative, forward slashes | e.g. `Modules/Sales/Sales.Services/Foo.cs` |
-| `line` / `lineEnd` | Integer (1-based), or `null` if unknown | Set both to same value for single-line remarks |
-| `suggestedFix` | Omit the field if remark is a `note` | |
-| `body` | GFM Markdown string | Used as the top-level PR review body |
-
----
-
-## CLI Import Commands
-
-After the agent saves both files, output a **ready-to-run commands block** in the chat so the user can import the review into their PR immediately.
-
-### GitHub — `gh` CLI
-
-```bash
-# Option A: Post the full Markdown report as a single PR comment
-gh pr comment {PR_NUMBER} --body-file docs/code-reviews/{issueId}-review_{timestamp}.md
-
-# Option B: Post as a formal PR review with inline file comments (uses GitHub Review API)
-# Requires: GH_TOKEN with repo scope, jq installed
-gh api repos/{OWNER}/{REPO}/pulls/{PR_NUMBER}/reviews \
-  --method POST \
-  --input docs/code-reviews/{issueId}-review_{timestamp}.json
-```
-
-> **Note for Option B:** The JSON payload's top-level structure matches the GitHub PR Review API.
-> GitHub inline comments (`remarks[]`) require `path` (= `file`), `line`, and `body`.
-> The `gh api --input` flag reads JSON directly from the file — no shell escaping needed.
-
-### Azure DevOps — REST API via `curl`
-
-```bash
-# Post the full Markdown report as a PR thread (top-level comment)
-curl -s -X POST \
-  "https://dev.azure.com/{ORG}/{PROJECT}/_apis/git/repositories/{REPO}/pullRequests/{PR_ID}/threads?api-version=7.1" \
-  -H "Content-Type: application/json" \
-  -H "Authorization: Bearer $AZURE_DEVOPS_TOKEN" \
-  --data-binary @- <<EOF
-{
-  "comments": [{"content": $(jq -Rs . < docs/code-reviews/{issueId}-review_{timestamp}.md), "commentType": 1}],
-  "status": 1
-}
-EOF
-
-# Post individual inline remarks from the JSON payload
-# (run once per remark that has a non-null 'file' and 'line')
-curl -s -X POST \
-  "https://dev.azure.com/{ORG}/{PROJECT}/_apis/git/repositories/{REPO}/pullRequests/{PR_ID}/threads?api-version=7.1" \
-  -H "Content-Type: application/json" \
-  -H "Authorization: Bearer $AZURE_DEVOPS_TOKEN" \
-  -d '{
-    "comments": [{"content": "{R1 body}", "commentType": 1}],
-    "threadContext": {
-      "filePath": "/{file}",
-      "rightFileStart": {"line": {line}, "offset": 1},
-      "rightFileEnd": {"line": {lineEnd}, "offset": 1}
-    },
-    "status": 1
-  }'
-```
-
-> The agent outputs **placeholder-filled** commands (with actual values substituted) so the user only needs to set environment variables (`OWNER`, `REPO`, `PR_NUMBER`, `AZURE_DEVOPS_TOKEN`) and run.
-
----
-
 ## Formatting Rules
 
 ### General
@@ -293,19 +180,6 @@ curl -s -X POST \
 - Range: `L42-L58`
 - If line number cannot be determined, use `(general)` and explain the scope in the description
 
-### GitHub PR Compatibility
-The report is designed to render correctly when pasted as a GitHub PR comment:
-- Tables use standard GitHub-flavored Markdown (GFM)
-- Emoji severities (🔴🟡🟢ℹ️) render natively
-- No HTML tags — pure Markdown only
-- Collapsible sections are NOT used (inconsistent support across platforms)
-- Code references use backtick-wrapped paths, not links (links break across forks)
-
-### Azure DevOps PR Compatibility
-- Azure DevOps supports GFM in PR comments
-- Same Markdown renders correctly in both platforms
-- Avoid Mermaid diagrams in the report (not supported in Azure DevOps PR comments)
-
 ## Edge Cases
 
 ### No Remarks Found