nightgauge

Re-review the current change with fresh-eyes critics that have no authoring context, each attacking it from a distinct lens (correctness, security, reuse/simplification, tests), then drive a fix loop until findings degrade to nitpicks. Use to harden a diff before opening or merging a PR, when asked for an adversarial/critical review, or from feature-dev/pr-create as an optional hardening gate.

Analyze an epic's sub-issues and recommend batch vs sequential pipeline processing strategy. Use before starting work on an epic to determine optimal execution approach.

Groom the open backlog so nothing dispatched is wasted work — re-assess every open issue against current main on five axes (validity, worth, verification completeness, security of the proposed approach, epic fit), adversarially verify every proposed mutation, then apply append-only corrections, evidence-cited closures, epic moves and a stale-sync sweep. Repo-scoped or workspace-wide with cross-repo epic batches. Dry-run by default; use after an epic completes, before a planning session, or on a monthly cadence.

Validate backlog issues are pipeline-ready before processing. Checks required labels, acceptance criteria quality, dependency cycles, and greenfield readiness. Use before starting pipeline runs on a new or unfamiliar repo.

Turn on an opt-in guardrail that blocks production-destructive shell commands (docker compose down -v, docker volume rm, kubectl delete, SQL DROP/TRUNCATE) for the rest of the working session. Use before touching a production droplet, database, or cluster — when you want a safety net while doing risky ops — and turn it off when done.

Triage a failing CI check that no issue exists for — reproduce it, observe the running system, then fix it, in that order. Refuses to propose a fix it could not reproduce, files a type:spike instead, and emits a triage record naming the hypotheses it ruled out and the observation that ruled each one out. Use when a named check, workflow run or workflow file is red and nothing is tracking it.

Reference for Nightgauge' own surfaces — the `nightgauge forge` abstraction, the `nightgauge` Go binary subcommands, and the @nightgauge/sdk API — with the gotchas that bite when you use them. Use when authoring or editing a skill, plugin command, automation, or script that shells out to `nightgauge`/`forge` or imports the SDK, or whenever you are about to reach for a bare `gh`/`glab` call and need the forge equivalent.

Display the effective Nightgauge configuration with source annotations, showing where each value comes from (default, global, project, or environment). Use when debugging config precedence or confirming which tier supplies a value.

Unified continuous improvement review — orchestrates all self-improvement mechanisms (skill assessments, health analysis, calibration, recommendations, feedback loops) into a periodic review cycle. Two modes — dogfood (internal, proposes code/skill/doc changes) and customer (external, proposes config and workflow adjustments). Use on a regular cadence (weekly/bi-weekly) or after significant pipeline changes.

Dependency Modernization Engine — safely identifies and updates outdated or vulnerable dependencies with compatibility analysis, breaking change detection, and staged rollout recommendations. Use when auditing or upgrading dependencies, or after a security advisory flags a vulnerable package.

Auto-generate and update documentation for public APIs. Detects undocumented functions, generates JSDoc/docstrings, identifies signature changes, and suggests README updates. Use after /feature-dev or standalone on any codebase.

Monitor Claude Code documentation for new features and changes relevant to the pipeline. Use when checking for Claude Code updates or when a new CC version is released.

Write narrative architecture documentation sections by reading source files and synthesizing accurate, validated content. Use when picking up a documentation issue that requires reading code and writing a structured doc section — without a PLAN.md approval cycle.

Answers one question with evidence instead of opinion — is the Nightgauge pipeline ready to take over development of a given workspace, replacing an interactive agent session? Evaluates twelve objective, mechanically-checkable gates across instrumentation, provisioning, safety and release readiness, and returns a go / no-go with the current value of every gate. Use before handing a workspace to autonomous mode, before a release decision, or whenever someone asks "are we ready to let the pipeline build this yet?"

DEPRECATED — thin wrapper that delegates to nightgauge-issue-audit --epic <N>. Use only when an existing slash-command caller or documented workflow still invokes epic-validate; new code should call /nightgauge:issue-audit directly.

Implement features following the approved PLAN.md and documented standards. Includes quality review against docs/ files. Use after /feature-planning or when implementing any planned feature.

Documentation-first feature planning. Read docs before code, propose an implementation approach, and write a plan file under .nightgauge/plans/ for approval. Use after /issue-pickup and before /feature-dev to produce the approved PLAN.md.

Validate feature implementation with integration/E2E tests and manual checklists. Use after /feature-dev to ensure the feature works end-to-end before creating PR.

Comprehensive codebase health assessment producing quantitative scores across 6 dimensions. Use when inheriting, auditing, or maintaining a codebase to understand its current state and prioritize improvements.

Cross-repository integration audit. Validates that client API calls match platform endpoints, auth flows are aligned, docs are current, and cross-repo dependencies are tracked. Run before major epic creation or after platform changes.

Deterministic post-creation gate that verifies every issue created by issue-create (or any creation flow) has the labels, board membership, body sections, sub-issue links, blockedBy edges, cross-repo references, and knowledge scaffold the pipeline depends on. Reports findings by severity and offers --fix / --fix-interactive auto-repair using existing Go binary primitives. Use after every issue-creation flow or as a CI safety net.

Create well-structured GitHub issues with SDLC metadata, project board sync, and optional parent/child linking. Use when filing a new issue or epic that the pipeline will pick up, so it lands board-ready with correct labels and links.

Claim a GitHub issue, extract requirements, and set up the development environment. Use at the start of any issue-based development work.

Analyze a raw GitHub issue and rewrite it with structured sections, acceptance criteria, and implementation guidance — making it immediately pipeline-ready. Use before /nightgauge-issue-pickup on sparse or unstructured issues.

Modernization plan generator that consumes health-check, security-audit, and test-scaffold outputs to produce a prioritized, phased roadmap. Use after running assessment skills to get an actionable improvement plan.

Deterministic codebase pattern mining subagent. Discovers existing implementation patterns, naming conventions, and structural organization before planning. Invoked by feature-planning as a Task subagent — not user-facing.

Analyze pipeline execution history for efficiency insights - compute token usage, stage performance, cost optimization, quality correlation, and trend metrics. Use anytime to identify improvement opportunities.

Comprehensive pipeline health analysis across 7 dimensions — token economics, cost health, stage effectiveness, model routing, reliability, self-improvement loop health, and pipeline velocity. Cross-references all telemetry data sources to generate actionable findings and auto-create improvement issues. Use periodically or after a batch of pipeline runs to assess overall pipeline health.

Triage recent pipeline failures end-to-end — distinguish FALSE alarms (work already merged) from real failures, classify the root cause against a known-failure taxonomy, and either apply the safe operational fix (re-queue, clear false pause/cooldown) or report a precise diagnosis. Use whenever the pipeline reports a failure or autonomous mode is paused.

Create a pull request with correct base/head, issue linkage, validation summary, and reviewer assignment. Use after /feature-validate to open the PR for an implemented issue.

Wait for PR reviews, address feedback, and merge. Completes the Issue-to-PR pipeline after /pr-create.

Universal pre-flight validation for pull requests. Run before submitting PRs to catch common issues like broken links, invalid syntax, and missing documentation. Works on any repository.

Comprehensive 8-dimension product quality audit across all Nightgauge repositories. Validates API alignment, epic lifecycle, documentation accuracy, feature parity, test coverage, security posture, dependency health, and CI/CD integrity. Run weekly or before major releases.

Bulk-sync existing repository issues to GitHub Project boards with proper field mappings. Use during onboarding or catch-up to sync dates from milestones and the Status field.

Manage the issue queue for sequential and batch pipeline processing. Add, list, remove, and clear queued issues. Supports epic expansion and label-based queuing. Use when staging which issues the pipeline processes next, or when clearing the queue.

Refactor vs rewrite decision analysis engine. Evaluates brownfield codebases across 8 dimensions to produce data-driven recommendations with confidence levels, risk/benefit matrices, and hybrid approach suggestions. Use when deciding whether to refactor or rewrite a legacy component before committing to a modernization path.

Draft user-facing fastlane release notes (iOS/macOS + Android) from a closed epic's sub-issue titles and bodies, sized to satisfy the store-deploy freshness gate. Use after an epic fully closes to produce a human-reviewed "what's new" draft before a store deploy — never auto-submitted.

Monitor Claude Code GitHub releases for new features and changes relevant to the pipeline. Classifies changes by type and scores pipeline relevance. Use after a Claude Code release or on a periodic cadence to surface pipeline-relevant changes.

Prime a new GitHub repository and project board for the Nightgauge SDLC pipeline. Creates standard labels, validates/creates project board fields, creates standard project board views (Backlog, Priority board, Team items, Roadmap, My items), links the repo to the project, and generates .nightgauge/config.yaml with all field IDs. Discovers and reuses existing projects (org-preferred) before creating new ones. Validates token scopes upfront. Idempotent — safe to re-run at any time. Use once when onboarding a fresh repository for the pipeline, or to fill gaps and refresh field IDs.

Analyze pipeline failures to identify root causes, recurring patterns, and actionable remediation steps using deterministic binary classification. Also records outcome data and lessons learned to the knowledge base when knowledge_path is available. Use after a batch run or periodically to understand why runs did not complete successfully.

Comprehensive codebase security assessment producing quantitative scores across 7 dimensions. Use when auditing a codebase for vulnerabilities, hardcoded secrets, OWASP Top 10 risks, weak cryptography, input validation gaps, authentication flaws, and misconfiguration.

Make any repository AI-ready with AGENTS.md, CLAUDE.md, and focused documentation. Use when setting up a new project or when a repository is missing AI configuration files.

Parse a spike artifact's YAML recommendations block and create follow-up GitHub issues idempotently. Runs as the final stage of the spike pipeline path, after pr-merge.

Generate comprehensive test suites with coverage analysis using parallel subagents. Supports Jest, Pytest, dotnet test, and Gradle. Use after /feature-dev or when any codebase needs better test coverage.

Analyze existing test coverage, identify critical untested paths, and generate characterization tests as a safety net before refactoring. Use before any modernization to pin current behavior.

Verify and update documentation to match current codebase - detect drift, deprecated references, and inconsistencies. Use when documentation may be stale or after significant codebase changes.

Drive a running UI through a critical user flow with the Playwright CLI, asserting state at each step and capturing screenshots/traces, to verify a change actually works in the browser (not just that tests pass). Use after a UI-affecting change in a UI-bearing repo (dashboard, flutter web, acme-site, acme-web), when asked to visually verify a fix, or from feature-validate when the diff touches frontend code.

Derive the next semantic version (major/minor/patch) and a Keep-a-Changelog entry from a closed epic's merged sub-issues, then write the pubspec.yaml version name and CHANGELOG.md so they stay in sync with the store release notes. Use after an epic fully closes, before a store deploy, to land the version bump via the normal PR flow — the build number stays store-anchored and nothing is auto-submitted.

Scaffold a multi-repo workspace manifest (.vscode/nightgauge-workspace.yaml) for the N:1 shared-project topology, where several member repositories share a single GitHub Project. Detects member repos under the parent folder, derives the shared project, generates the manifest with repositories/routing/epic blocks, and verifies the result via `workspace sync-payload`. Idempotent — re-running refreshes the manifest without duplicating entries. Use once when onboarding a parent folder that groups multiple pipeline repos, or to repair an empty Repositories view.