planlisted

<objective> Wizard: scans codebase, proposes metric/guard/agent config, writes `program.md` run spec. Also runs cProfile on file path to surface bottlenecks before prompting for optimization goal. NOT for: running experiments (use `/research:run`); methodology validation (use `/research:judge`); full pipeline from goal to result (use `/research:sweep`); benchmarking or microbenchmark design (use `foundry:perf-optimizer` — plan's scope is ML metric optimization loops, not raw latency/throughput benchmarking). </objective> <workflow> <!-- Agent resolution: see _RESEARCH_SHARED/agent-resolution.md --> ## Agent Resolution **Environment precondition** — `CLAUDE_PLUGIN_ROOT` is set automatically when this skill runs via the plugin manager. Fallback `plugins/research` resolves only from project root. If neither resolves (bare `.claude/` copy invoked from a subdirectory), bin/ scripts return empty strings silently. **bin/ scripts this skill depends on** (deployed inside `${CLAUDE_PLUGIN_ROOT}/bin/`): `resolve_shared.py`, `make_run_dir.py`, `health_monitor_start.py`. Each call below is followed by an explicit empty-result guard — silent failure surfaces as a fail-fast error, never as an empty-string path. ```bash _RESEARCH_SHARED=$(python "${CLAUDE_PLUGIN_ROOT:-plugins/research}/bin/resolve_shared.py" 2>/dev/null) # timeout: 5000 [ -z "$_RESEARCH_SHARED" ] && { echo "! Plugin path resolution failed — ensure research plugin installed and CLAUDE_PLUGIN_ROOT set, or invoke /res