Sprint Plan¶

Planning Assumptions¶

Sprint length: 2 weeks
Capacity assumption: 1 full-stack engineer plus light design and review support
Estimation model: story points with a rough engineering-day range
Priority rule: security and prompt-quality guardrails come before analytics and platform expansion

Sprint 1: V1.1 Security and Prompt Quality Foundation¶

Goal: harden the browser extension without rebuilding the MVP baseline.

#10 Sensitive Data Detection (Security Guardrail) — 5 points — 2 to 3 days
#11 Prompt Quality Engine v2 — 8 points — 4 to 5 days
#13 SEO-Friendly Project Structure and Discoverability Docs — 3 points — 1 to 2 days

Total: 16 points

Expected outcome: - prompt submission can warn on likely secrets before AI send - prompt scoring becomes more structured and more explainable - docs and README improve discoverability for developer search traffic

Sprint 2: Coaching Quality and Analytics Foundation¶

Goal: make coaching more actionable and start capturing local learning signals.

#12 Prompt Linter (ESLint for AI Prompts) — 5 points — 2 to 3 days
#14 Prompt Session Tracking — 5 points — 2 to 3 days
#15 Session Summary — 5 points — 2 to 3 days
#17 Role-Based Coaching — 5 points — 2 to 3 days

Total: 20 points

Expected outcome: - users get lint-style prompt feedback before send - local prompt event history exists for future insight features - daily summary and role-aware coaching become visible product value

Sprint 3: Trend Visibility and IDE Foundation¶

Goal: show progress over time and open the path into IDE workflows.

#16 Skill Trend Graph — 5 points — 2 to 3 days
#18 VS Code Extension Scaffold — 8 points — 4 to 5 days
#19 Debugging-First IDE Templates — 5 points — 2 to 3 days

Total: 18 points

Expected outcome: - users can see trend lines for prompt quality and warning frequency - VS Code companion exists with shared prompt-analysis foundations - debugging templates accelerate real developer workflows

Sprint 4: Review Workflow and Policy Foundation¶

Goal: extend coaching into review flows and organizational guardrails.

#20 PR Coaching Hooks — 8 points — 4 to 5 days
#23 Policy Packs — 8 points — 4 to 5 days

Total: 16 points

Expected outcome: - teams can start reviewing AI-assisted code more responsibly - organizations gain early policy controls for risky AI usage

Sprint 5: Team Insights¶

Goal: provide manager-facing visibility without losing privacy discipline.

#21 Team Dashboard — 8 points — 4 to 5 days
#22 Manager Insights — 5 points — 2 to 3 days

Total: 13 points

Expected outcome: - managers can see aggregate prompt-quality and warning trends - coaching reports highlight weak areas and improvement opportunities

Current Execution Order¶

#10 Sensitive Data Detection
#11 Prompt Quality Engine v2
#12 Prompt Linter
#14 Prompt Session Tracking
#15 Session Summary

This ordering keeps the product safer first, then makes coaching smarter, then adds learning analytics on top of better-quality signals.