| name | project-overview-assessment |
| description | High-level assessment for the HealthGuard project. Triggers on keywords: project structure, project overview, tong quan, review architecture, assess structure. Uses MASTER_INDEX to optimize context. |
| risk | safe |
| source | custom |
| date_added | 2026-03-03 |
| date_updated | 2026-03-04 |
Skill: Project Overview Assessment (TongQuan)
Purpose
Evaluate the holistic structure and progress of the HealthGuard project to ensure it aligns with the SRS, follows industry-standard architecture, and stays on track with JIRA sprint tasks (Epics/Stories).
When to Use This Skill
Automatically activates when the user wants to:
- Evaluate the overall project or a specific large section
- Verify if the existing codebase structure reflects the SRS correctly
- Review progress against JIRA sprints (Epics & Stories)
- Report the project status to stakeholders
Context Loading Protocol (MANDATORY)
[!IMPORTANT]
The Agent MUST strictly follow this 3-tier loading protocol to optimize token limits and prevent getting lost in context.
Tier 1: Navigation (ALWAYS read first)
- Read
PM_REVIEW/MASTER_INDEX.md — The overall GPS map of the project.
- Read
PM_REVIEW/Resources/SRS_INDEX.md — System-level context (architecture, features, thresholds).
- Determine the assessment scope (Admin / Mobile / or both).
Tier 2: Prepare Integrated Skills
- CRITICAL: Before evaluating, read the internal bundled skills under
skills/ to inherit their high-level structural rules:
skills/architect-review/SKILL.md
skills/software-architecture/SKILL.md
skills/backend-architect/SKILL.md
skills/architecture-patterns/SKILL.md
Tier 3: Structure (Read based on scope)
- Read the corresponding
Project_Structure.md (Admin or Mobile).
Tier 4: Summaries (Read ONLY relevant summaries)
- Read related summary files from the
summaries/ folder.
⛔ WHAT NOT TO DO
- ❌ DO NOT read the full SRS document — use the summary files instead. If more detail is needed, read the specific Use Case (UC) file in
PM_REVIEW/Resources/UC/.
- ❌ DO NOT read the full JIRA CSV — use
PM_REVIEW/Resources/TASK/JIRA/README.md (JIRA Index) to quickly locate Epics/Stories, then read only relevant rows from the CSV.
- ❌ DO NOT read the entire source code — only scan the folder structure for overviews.
Evaluation Process (Progressive Deepening)
Step 1: Complete the Context Loading Protocol.
Step 2: Source Code Scan (Surface Level). Browse folder structure without reading file contents, comparing against Project_Structure.md.
Step 3: Evaluate against 6 Criteria: SRS Compliance, Architecture, Consistency, Progress vs JIRA, Code Quality, Security. READ references/evaluation-criteria.md for specific check details.
Step 4: Cross-check with JIRA Stories from the JIRA Index.
Output File Protocol (MANDATORY)
File Naming Convention
The review file MUST follow this naming pattern:
TONGQUAN_{PROJECT}_review.md
- PROJECT: Project name, UPPERCASE (e.g.,
ADMIN, MOBILE)
- If reviewing both projects → create 2 separate files or
TONGQUAN_ALL_review.md
Output Location
- Admin project →
PM_REVIEW/REVIEW_ADMIN/TONGQUAN_ADMIN_review.md
- Mobile project →
PM_REVIEW/REVIEW_MOBILE/TONGQUAN_MOBILE_review.md
Examples
| User Request | File Output |
|---|
| "Overview assessment for Admin project" | REVIEW_ADMIN/TONGQUAN_ADMIN_review.md |
| "Overview assessment for Mobile" | REVIEW_MOBILE/TONGQUAN_MOBILE_review.md |
Re-review Protocol (2nd review and beyond)
[!IMPORTANT]
Before starting a review, the AI MUST check if a previous review file already exists. If found → perform comparison.
Step 1: Find previous review file
- Determine the filename using the Output File Protocol above.
- Check if the file exists in
REVIEW_ADMIN/ or REVIEW_MOBILE/.
Step 2: Read previous review file (if exists)
- Read the old review file and extract:
- Old score: Total score + score per criterion.
- Old weaknesses: List of all weaknesses.
- Old recommendations: List of action recommendations.
- Old review date: From the "Thông tin chung" section.
- Old review count: From the "Thông tin chung" section (if present).
Step 3: Perform new review as normal
- Execute the review following the Evaluation Process above (Step 1–4).
Step 4: Compare and evaluate changes
- Compare new results with data extracted from the old file:
- Score increase/decrease per criterion.
- Weaknesses that have been fixed (present in old, absent in new).
- Weaknesses that still exist (present in both old and new).
- Weaknesses that are newly introduced (absent in old, present in new).
Step 5: Overwrite old file
- OVERWRITE the old review file with the complete new report, including the "🔄 SO SÁNH VỚI LẦN ĐÁNH GIÁ TRƯỚC" section (per template
references/report-template.md).
- Increment the Lần đánh giá counter by 1.
Important Notes
- If NO previous file found → this is the first review → do NOT add comparison section, set
Lần đánh giá: 1.
- If previous file found → MUST add comparison section and increment
Lần đánh giá.
Output Formatting
MANDATORY: You must use the Vietnamese markdown reporting template located at references/report-template.md. You must not deviate from this template.
After Review: Update MASTER_INDEX (MANDATORY)
When the overview review concludes, you MUST update MASTER_INDEX.md:
- Set
Review Status → ✅ Done (for each module assessed)
- Set
Score → XX/100 (if individual module scores are given)
- Set
Quality → Determined by score (see DanhGiaChiTiet/references/evaluation-criteria.md → Score Classification):
- 76–100 → ✅ Pass
- 51–75 → ⚠️ Needs Fix
- 0–50 → ❌ Fail
- Set
Review File → Link to the overview review file (e.g., [View](REVIEW_ADMIN/TONGQUAN_ADMIN_review.md))
- Set
Last Review → Current date (ISO format)
Edge Cases
| Scenario | Action |
|---|
| Summary file does not exist for a module | Note as "Missing summary" in the report, skip detailed assessment for that module |
| Source code folder is empty / not built | Mark as "⬜ Not built" in report, score = N/A |
| MASTER_INDEX row missing for a module | Add the row with available info |
| Previous review file is corrupted/unreadable | Treat as first review, set Lần đánh giá: 1 |
Reference Documents
| Name | Path | When to read |
|---|
| MASTER INDEX | PM_REVIEW/MASTER_INDEX.md | ALWAYS |
| SRS Index | PM_REVIEW/Resources/SRS_INDEX.md | ALWAYS — System-level context |
| Admin Structure | PM_REVIEW/REVIEW_ADMIN/Project_Structure.md | When reviewing Admin |
| Mobile Structure | PM_REVIEW/REVIEW_MOBILE/Project_Structure.md | When reviewing Mobile |
| Admin Summaries | PM_REVIEW/REVIEW_ADMIN/summaries/*.md | Based on module |
| Mobile Summaries | PM_REVIEW/REVIEW_MOBILE/summaries/*.md | Based on module |
| DB Summary | PM_REVIEW/SQL SCRIPTS/README.md | When reviewing database design |
| Use Cases (UC) | PM_REVIEW/Resources/UC/**/*.md | When explicit detail is missing from summaries |
| JIRA Index | PM_REVIEW/Resources/TASK/JIRA/README.md | ALWAYS — Quick lookup for Epics/Stories |
| JIRA Full CSV | PM_REVIEW/Resources/TASK/JIRA/JIRA_IMPORT_ALL.csv | Only when need full Story details |