wieldlisted

# Wield — Functional QA (You wield the blade the way a user would.) Read `${CLAUDE_PLUGIN_ROOT}/references/creed.md`, `${CLAUDE_PLUGIN_ROOT}/references/stacks.md`, and `${CLAUDE_PLUGIN_ROOT}/references/dispatch.md` first. Log: `bash ${CLAUDE_PLUGIN_ROOT}/scripts/ledger.sh append temper <slug> wield STARTED -` ## Process 1. **Detect stack + surface:** `stack-detect.sh`. Playbook from this skill's `references/` dir by `stack=`: `ts.md`, `python.md`, `go.md`, `java.md`, `rust.md` (any web UI can additionally use Playwright via npx per ts.md). Web UIs and services need a runnable target — get the run command + URL from spec.md / STATE.md or ask. 2. **Pick the tier** (AskUserQuestion, default Standard): - **Quick** — fix-worthy findings: Critical + High only - **Standard** — + Medium - **Exhaustive** — + Low/cosmetic 3. **Derive the flow list** from spec.md success criteria (pipeline mode) or from the user (standalone). Each flow: steps, expected outcome, edge and error variants. No invented requirements — flows trace to the spec. **Persona mode — automatic when `$SMITHY_MEM/personas/` exists** (created by `/smithy:commission`; offer to run it if absent on a multi-role app): - Derive flows PER persona from each persona's jobs-to-be-done, executed within that persona's permission boundary (their credentials/role). - Add **cross-persona checks**: every `CANNOT` in a persona file becomes a test that the action/data is actually d