← ClaudeAtlas

commissionlisted

Generate project test personas from real user roles (evidence + interview); powers per-persona QA and guild judgment. Triggers: 'commission', 'define the users'.
dyasnurhakim/smithy-claude · ★ 0 · Testing & QA · score 67
Install: claude install-skill dyasnurhakim/smithy-claude
# Commission — Project Personas (A commission tells the smith who the work is FOR. You cannot judge a blade without knowing whose hand it fits.) Read `${CLAUDE_PLUGIN_ROOT}/references/creed.md` and `${CLAUDE_PLUGIN_ROOT}/references/envelope.md` first. Resolve memory first: `export SMITHY_MEM="$(bash ${CLAUDE_PLUGIN_ROOT}/scripts/paths.sh mem)"` — every smithy path below is relative to it, and it need NOT be inside the repo. If that dir does not exist, bootstrap per `${CLAUDE_PLUGIN_ROOT}/references/memory.md` § Location. Log: `bash ${CLAUDE_PLUGIN_ROOT}/scripts/ledger.sh append commission <slug-or-'-'> personas STARTED -` Output: one file per role at `$SMITHY_MEM/personas/<role-slug>.md`. These are PROJECT personas (specific real users of THIS system) — distinct from the plugin's generic masters/patrons, which they complement. ## Process 1. **Discover the roles — evidence first, questions second.** - Read `$SMITHY_MEM/jobs/*/spec.md`, README, and product docs for named user types. - Grep the code for role definitions: `role`, `permission`, `is_admin`, enums like `Role.`, auth middleware, route guards, RBAC tables, seed data. Cite file:line for every role found. - Present the discovered list; ask the user what's missing or wrong ("who else uses this system? anyone non-obvious — internal admins, API consumers, auditors?"). Never invent a role the evidence or the user didn't give you. 2. **Interview per role** (AskUserQuestion, batche