Developer · code review + platform builds

Developer

Reviews all code changes from agents and humans. Enforces codebase conventions, security, tutor instruction layer integrity, deployment rules. Owns Track B bug fixes + platform code.

Not in current cascade

Agent file.claude/agents/developer.md @ 30e888e9
Team membershipSpecialized direct-instance roster
ToolsRead · Write · Edit · Bash · Grep · Glob
Catalog sectionAGENT-ROLES-AND-RESPONSIBILITIES-CATALOG §3.3
Existence statusexists

Summary

Reviews all code changes from agents and humans. Enforces codebase conventions, security, tutor instruction layer integrity, and deployment rules. Also owns Track B bug fixes and platform code.

Invoke before any commit touching `platform/` or tutor layers. Three-skill review pattern (mandatory per CLAUDE.md SO #16): architecture-review + security-review + code-review. All three sections must appear in the review output.

Distinguishing characteristicsDifferent from QA because: developer reviews code (architecture · security · code quality), QA verifies runtime behavior. Different from Validator because: developer scope is platform code, Validator scope is curriculum content. Different from code-pm because: developer reviews + writes code (Read/Write/Edit/Bash), code-pm is read-only governance verifier.

Research portfolio

Placeholder reasonNot in current Tier 2 cascade scope (cascade prioritized 8 classes excluding Developer). Developer's existing 3-skill review framework (architecture-review · security-review · code-review) covers core scope. Future M-1 brief amendment may add Developer to cascade extension; Researcher structural eval per catalog §4.4 routing.

Significant project contributionsCumulative S40+ → Cumulative S40+

Cumulative S40+ · ongoing · platform/* code review + Track B bug fixes
Standing review responsibility for any commit touching `platform/src/**`, `platform/scripts/**`, tutor layers (`tutor-base.ts`, `tool-reminders.ts`, `tutor-prompt.ts`, `mcp-remote-instructions.ts`). Three-skill review pattern enforced per CLAUDE.md SO #16.

Learningsauthored + cross-cutting

Cross-cutting applied

PRINCIPLE 16 — internal identifiers never render to end users (S60 origin)
No snake_case / kebab-case identifier-style / camelCase programmatic-key token renders as visible text on student- or prospect-facing surfaces. Pattern-match subset of PRINCIPLE 1.
Three-skill Developer review (CLAUDE.md SO #16)
Platform code changes require Developer invocation with all three review skills: architecture-review, security-review, and code-review. All three sections must appear in the review output.

Skills + hooksused + constraining

Skills used (always-loaded)

Quality protocol

code-review

Conventions · correctness · performance · render paths · tutor layer integrity. 0-100 confidence scoring.

DCAR-based

architecture-review

Server/client boundaries · DAL pattern · Supabase client selection · coupling analysis (Next.js App Router + Supabase + Vercel).

OWASP Top 10:2025

security-review

OWASP Top 10:2025 mapped to Next.js + Supabase + Vercel stack. 20+ security checks with confidence scoring.

Skills task-match-loaded

Token consumption

design-system

AI Uni visual identity · brand tokens · component patterns. Consumed when implementing visual-token decisions.

Stack-specific

supabase

Supabase Database · Auth · Edge Functions · Realtime · Storage · RLS. Triggered when code imports supabase-js / @supabase/ssr.

Hooks constraining

Tool-scope · agent file frontmatter
Read · Write · Edit · Bash · Grep · Glob — full execution authority for platform code. Distinguished from review-only agents.
Deploy-review doc required (CLAUDE.md SO #8)
Produce `docs/reviews/DEPLOY-REVIEW-S{N}-TASK{M}.md` per `docs/GOVERNANCE.md` section 1. Run `bash scripts/validate-signoff-ready.sh`. No review doc = no deploy request.
Three-skill review pattern (CLAUDE.md SO #16)
Platform code changes require all three review skills (architecture + security + code). All three sections must appear in review output.

Last updated · refresh details

Profile auto-regenerated2026-05-06T12:50:00Z
Refresh cadenceDaily midnight UTC · per-session-end · on-demand button (admin-tier optional)
agent file.claude/agents/developer.md @ 30e888e9
corpusnot yet authored
catalogdocs/agent-knowledge/AGENT-ROLES-AND-RESPONSIBILITIES-CATALOG.md @ d7d1aee8+241affb8

Refresh strategy per HANDOFF v4 §12 — profile auto-regenerates from agent-file + corpus + executions + git log.

Cross-references from the relationship graph

Edges sourced from cross-refs/matrix.json · also browsable on the org chart.

Cross-references 5 skills · 2 hooks

Skills · always-loaded

Code-Review
Conventions · correctness · performance · render paths · tutor-layer integrity.
Architecture-Review
DCAR-based — server/client boundaries · DAL pattern · Supabase client selection.
Security-Review
OWASP Top 10:2025 mapped to Next.js + Supabase + Vercel stack.

Skills · task-match-loaded

Design-System
Token consumption when implementing visual-token decisions.
Supabase
Supabase product family — Database · Auth · Edge Functions · RLS.

Hooks constraining

tool-scope-frontmatter · hard
Read · Write · Edit · Bash · Grep · Glob
check-deploy-review · hard
platform/* changes require deploy-review doc