edc-auditlisted

# Audit Code Quality Analyze code quality and maintainability using the `edc-context/` files as the baseline for what the code SHOULD look like. This is a read-only scoped analysis pass: audit workers must not mutate source code, branch state, generated context, or canonical reports outside their assigned report artifact. **Clean Slate Rule:** When spawning subagents for analysis, they MUST be fresh agents with NO access to the current conversation context. This prevents bias from user discussion influencing findings. See `edc-build-impl` for rationale. ## Required references Read these bundled references before producing findings: 1. `references/scope-and-standards.md` — scope discipline, standards sources, evidence rules, and tooling-skip policy 2. `references/smell-baseline.md` — Fowler smell baseline; use as heuristics and label baseline-only findings as possible smell 3. `references/quality-checks.md` — concrete code-quality checks for maintainability, correctness smells, error handling, tests, type/contracts, and simplicity 4. `references/reporting.md` — canonical report targets and finding format The runtime embeds these files below this `SKILL.md` for orchestrated audit runs. If invoking the skill manually and the references are not embedded, read them from this skill directory before continuing. ## Prerequisites This skill usually runs inside an audit subprocess spawned by `plugins/edc/scripts/edc-audit.sh` or a build-time audit orchestration step. The orches