Pharn Dev Regress — Development skill for Claude Code
Detect regressions OUTSIDE the just-built feature: re-run the existing deterministic suite (npm run check's gates) at the pre-build BASELINE and at HEAD, and flag any gate that flipped pass→fail.
How to install Pharn Dev Regress
Installs to ~/.claude/commands/pharn-dev-pharn-oss-pharn-dev-regress.md
mkdir -p ~/.claude/commands && curl -fsSL https://raw.githubusercontent.com/pharn-dev/pharn-oss/HEAD/.claude/commands/pharn-dev-regress.md -o ~/.claude/commands/pharn-dev-pharn-oss-pharn-dev-regress.md Restart Claude Code, or start a new session, for it to be picked up.
What Pharn Dev Regress does
description: "Detect regressions OUTSIDE the just-built feature: re-run the existing deterministic suite (npm run check's gates) at the pre-build BASELINE and at HEAD, and flag any gate that flipped pass→fail. The verdict is a deterministic exit-code comparison (pharn/floor/check-regress.mjs) — ZERO LLM judgment in its core. Emits regression-report.json (machine) + REGRESSION.md (human). FLOOR verdict; ADVISORY orchestration." kind: pharn-owned trust: trusted model_tier: sonnet reads: ["phar
Alternatives in Development
- Om Tidy — Self-maintenance pass — acts on every hygiene flag: archives completed work, groups loose clusters, splits ove 4.6k ★
- Battle — Head-to-head graded battle card between two sites — eight dimensions, bar-by-bar, verdict line 3.6k ★
- Claude Code Project Boundary — Claude Code plugin: scope-aware protection — allows destructive ops (rm, mv, chmod) within project, blocks the 86 ★
Full documentation available on GitHub
View Source RepositoryRelated Skills
Pharn Dev Eval
Run a capability's eval LIVE via claude -p N times into isolated runs/, then COUNT structural pass/fail across
Init Project Doctor
Detect and repair baseline drift — missing or invalid .claude/workflows.jsonl, schema violations, four-way Art
Eval Prompts
Run a golden-set prompt eval suite against the pinned baseline; fail on any regression beyond threshold.
Pharn Dev Ship
Run PHARN's build loop in order so the human need not re-type or memorize it: /pharn-dev-plan → human approve
Bobby Performance
Performance benchmarking — measure and detect regressions
Head To Head
Head-to-head — measure ripwire against a baseline, honestly
Related Agents
Headroom Auditor
Runs before any treatment is built or believed. Given a task suite, a baseline and a proposed treatment, it es
Test Suite Analyzer
Detect one package's test suites for a Stop test hook - stack, runners, suite kind, hook command with fail-fas
Render QA
Renders a built artboard, records deterministic visual/GIF evidence, and applies render, contrast, seam, motio