dyasnurhakim
UserA full dev workflow plugin for Claude Code: research → plan → implement (TDD optional) → multi-persona review → test. It asks instead of assuming, remembers progress across sessions, and blocks destructive commands until you approve them.
Categories
Indexed Skills (20)
anneal
Debugging: reproduce → read-only RCA → approved minimal fix + regression test (always TDD). Triggers: 'anneal', 'debug', 'why is this broken'.
assay
Research → spec: explores the codebase, turns every assumption into a question or recommendation. Triggers: 'assay', 'research this', 'write a spec'.
blueprint
Spec → verify-annotated plan + self-contained task briefs, persona pass, parallel batch markers. Triggers: 'blueprint', 'plan this', 'break into tasks'.
burnish
Design review & improvement: screenshot the live UI, judge vs DESIGN.md, gated surgical fixes with before/after proof. Triggers: 'burnish', 'polish the UI'.
calibrate
View/edit smithy config at GLOBAL (all projects) or PROJECT scope — model+effort routing, TDD default, gates, review panel, where memory lives. Probes model availability before writing. Triggers: 'calibrate', 'smithy config'.
commission
Generate project test personas from real user roles (evidence + interview); powers per-persona QA and guild judgment. Triggers: 'commission', 'define the users'.
forge
Execute the approved plan task-by-task (forger or jigsmith), review after every task, parallel batches in worktrees. Triggers: 'forge', 'implement the plan'.
guild
Production-readiness panel: parallel persona reviewers (masters=craft, patrons=experience) → one PRODUCTION_READY|NOT_READY verdict. Triggers: 'guild', 'ready to ship?'.
handover
Evidence-cited session handoff; next session resumes with zero re-discovery. Triggers: 'handover', 'handoff', 'save session'.
hone
Performance: baseline-first benchmarks (median of ≥3), profiler-evidenced hot spots, recommendations only. Triggers: 'hone', 'why is it slow', 'benchmark'.
inspect
Two-verdict code review (spec compliance + quality); findings need evidence, severity rationale, confidence. Triggers: 'inspect', 'review this change'.
jig
TDD implementation: RED→GREEN→REFACTOR per requirement with verbatim failing-test evidence. Triggers: 'jig', 'TDD this', 'test-first', bug fixes.
pattern
Design creation: subject-grounded direction with HTML previews, tokens, states, voice → the project DESIGN.md. Triggers: 'pattern', 'design system', 'make it look good'.
proof
Stress/load test a running service against user-set thresholds (never invented). Triggers: 'proof', 'stress test', 'load test'.
ring-test
Unit tests per the stack playbook (write missing, run, flag flakes). Triggers: 'ring-test', 'unit test this'.
smithy
Full pipeline orchestrator (research→plan→implement→review→test) with approval gates and ledger resume. Triggers: 'run smithy', 'build end to end', 'full pipeline', resume smithy work.
strike
One-shot fix lane: small known changes without a spec — lightweight plan → one confirmation → forge (no TDD) → targeted tests → one report. Triggers: 'strike', 'quick fix', 'fix these items', 'small change', review/QA finding fixups.
temper
Testing umbrella: runs ring-test/wield/proof/hone, one READY|NOT READY verdict. Triggers: 'temper', 'test everything'.
using-smithy
Skill router: when to use which smithy skill, priority rules, red-flag rationalizations. Injected as a digest each session; invoke for the full router.
wield
Functional QA as a user: mandatory screenshots, 0-100 health score, persona flows, severity tiers. Triggers: 'wield', 'QA this', 'does it work'.
Bio shown is the top-scored skill's repo description as a fallback — real GitHub bios land in a future update.