Fps Vod Review Automation

FPS VOD Review Automation — Production Skill

Overview

This skill provides evidence-backed FPS performance analysis and coaching through a 6-step harness that combines real-time data aggregation, recognized coaching methods, and academic research into structured, risk-disclosed recommendations.

When This Skill Triggers

Use this skill when the user mentions:

• FPS game analysis, VOD review, demo analysis
• Performance coaching, improvement plans
• Aim training, mechanical skill assessment
• Tactical analysis, positioning, utility usage
• CS2, Valorant, Overwatch 2, Apex Legends, Rainbow Six Siege
• "How to improve at [FPS game]", "Review my gameplay"
• FPS coaching, skill assessment, rank improvement
• Mistake analysis, drill prescriptions, practice plans

Quick Start

/fps-vod-review-automation "Analyze my CS2 gameplay and tell me how to reach Global"

Architecture

SKILL.md (Entry Point)
    │
    ├── config/          # Configuration management
    ├── src/             # Core Python package
    ├── skills/          # Sub-skill definitions
    ├── scripts/         # Automation utilities
    ├── references/      # Domain knowledge & prompts
    ├── assets/          # Templates & schemas
    └── tools/           # External tools

Sub-Skills Registry

| Skill ID | File | Purpose | Phase | |----------|------|---------|-------| | sub-gather-requirements | skills/sub-gather-requirements.md | Intake & requirements gathering | 1 | | sub-evidence-collector | skills/sub-evidence-collector.md | Data fetching & evidence collection | 2 | | sub-core-analysis | skills/sub-core-analysis.md | VOD review & performance analysis | 3 | | sub-knowledge-updater | skills/sub-knowledge-updater.md | Academic evidence retrieval | 4 | | sub-advisor | skills/sub-advisor.md | Synthesis & recommendations | 5 |

Execution Flow

1. Pre-Flight: Language Detection (vi/en)
    ↓
2. Sub-Skill 1: Gather Requirements
    ↓
3. Sub-Skill 2: Evidence Collection
    ↓
4. Sub-Skill 3: Core Analysis
    ↓
5. Sub-Skill 4: Knowledge Update
    ↓
6. Sub-Skill 5: Advisory Synthesis
    ↓
7. Quality Gates (U1-U6, G1-G4)
    ↓
8. Output Delivery

Tools Available

• WebSearch — Live domain data and academic sources
• WebFetch — Scrape authoritative pages
• Read — Access SECOND-KNOWLEDGE-BRAIN.md and references
• Write — Update knowledge base
• Bash — Execute pipeline scripts
• Skill — Invoke sub-skills sequentially

Hooks System

See references/hooks.md for available lifecycle hooks:

• pre-analysis — Before step 1
• post-evidence — After step 2
• post-analysis — After step 3
• pre-advisory — Before step 5
• post-quality-gate — After all gates

Configuration

Environment variables (optional):

• FPS_VOD_ROOT — Project root directory
• FPS_VOD_LOG_LEVEL — Logging level (default: INFO)
• FPS_VOD_LANGUAGE — Default language (en/vi)
• FPS_VOD_STRICT_MODE — Enable strict quality gates (default: true)

Quality Standards

This skill enforces 10 quality gates:

• U1-U6: Universal gates (sources, disclosure, evidence hierarchy, language, template, traceability)
• G1-G4: Domain gates (round review, mistake taxonomy, drills, coaching science)

Output Format

See assets/output-template.md for the complete output structure.

Knowledge Base

The skill maintains SECOND-KNOWLEDGE-BRAIN.md with:

• Core coaching methods
• 15+ DOI-cited academic papers
• Authoritative sources
• SOTA techniques
• Self-update protocol

Dependencies

Python packages (see requirements.txt):

• pydantic, yaml, structlog, requests, tenacity

Testing

# Run all tests
pytest tests/ -v

# Run validators
python tools/validate_project.py
python tools/run_test_scenarios.py

Contributing

See CONTRIBUTING.md for guidelines.

License

MIT License — see LICENSE file.

---

Version: 1.0.0 Last Updated: 2026-07-20 Status: Production Ready

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: dungnotnull
• Source: dungnotnull/fps-vod-review-automation-agent-skill
• License: MIT

Install and usage instructions live in the source repository linked above.