Skip to content

nickgallick/perlantir-fleet / workspace-forge/skills/developer-patterns

developer-patterns

Unverified

Tracks patterns, blind spots, and habits of each developer agent. Calibrates reviews to catch their specific recurring mistakes faster.

Developer Pattern Memory

Purpose

After every review, log what was found. Over time, build a profile of each developer's strengths and weaknesses. Use this to front-load reviews — check their known blind spots FIRST, then do the full checklist.

Maks Profile

Recurring Issues (update after every review)

DateProjectIssueSeverityCategoryTimes Seen
2026-03-21MathMindNo COPPA age-gate or parental consent flow for children's math appP0COPPA/Compliance1
2026-03-21MathMindMissing COPPA-compliant privacy policy pageP2COPPA/Privacy1
2026-03-21MathMindHardcoded API keys in client-side codeP2Code Quality1
2026-03-21MathMindMissing input validation on user-facing formsP2Code Quality1
2026-03-21MathMindInconsistent error handling patterns across componentsP2Code Quality1
2026-03-21MathMindMissing alt text on educational imagesP2Accessibility1
2026-03-21MathMindTouch targets below 44px minimum for children's appP2Accessibility1
2026-03-21MathMindColor-only indicators for correct/incorrect answersP2Accessibility1
2026-03-21MathMindUnhandled promise rejections in async quiz logicP2Error Handling1
2026-03-21MathMindMissing error boundary around quiz component treeP2Error Handling1
2026-03-21MathMindUnoptimized image assets causing slow load on mobileP2Performance1
2026-03-21MathMindLayout broken on small screens (< 375px)P2Responsive1

Blind Spot Summary (update when patterns emerge)

After 1 review:

  • Top 3 most frequent mistake categories: Accessibility (3), Code Quality (3), Error Handling (2)
  • Things Maks consistently gets right: Component architecture, TypeScript types, Tailwind usage
  • Things Maks consistently misses: Accessibility details, compliance requirements (COPPA), error boundaries
  • Improving areas: [TRACK — need more data]
  • Degrading areas: [TRACK — need more data]

Review Calibration

Based on pattern data, prioritize these checks FIRST when reviewing Maks's code:

  1. Compliance/regulatory requirements (COPPA, GDPR, ADA) — especially for apps targeting specific demographics
  2. Accessibility — alt text, touch targets, color-only indicators
  3. Error handling — error boundaries, unhandled promises, Supabase .error checks
  4. Input validation at trust boundaries
  5. [FILL after more data]

Other Developer Profiles

(Add sections here if other agents or humans submit code for review)

How to Update This Skill

After EVERY review:

  1. Open this file
  2. Add a row to the Recurring Issues table for each issue found
  3. Increment "Times Seen" if it's a repeat pattern
  4. After every 5th review, re-evaluate the Blind Spot Summary and Review Calibration
  5. If a previously recurring issue stops appearing for 3+ consecutive reviews, move it to "Improving areas"

Pattern Metrics

  • Total reviews completed: 1
  • Total issues found: 12
  • Average issues per review: 12
  • Most common P0 category: COPPA/Compliance
  • Most common P1 category: [UPDATE after more data]
  • Reviews since last Maks blind spot repeat: 0

Changelog

  • 2026-03-21: Populated with MathMind review data (12 issues across 6 categories)
  • 2026-03-20: Initial skill — pattern tracking begins
reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis.Stars: 464kUpdated Claude CodeCodexVerifiedInstall
reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis.Stars: 464kUpdated Claude CodeCodexVerifiedInstall
diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness.Stars: 318kUpdated Claude CodeCodexUnverifiedInstall
dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependenciesStars: 318kUpdated Claude CodeCodexUnverifiedInstall