sonash-v0

Structured discovery-first planning for complex tasks. Ask exhaustive categorized questions, build a standalone decision record, then produce a step-by-step implementation plan with audit checkpoints and convergence-loop verification of diagnosis and plan claims for user approval before any code is written.

Multi-agent research engine that decomposes questions, dispatches parallel searcher agents, synthesizes findings with citations and confidence levels, runs mandatory contrarian/OTB challenges, gap-pursuit verification, and cross-model verification via Gemini CLI, and produces structured output with downstream adapters, research index, and management commands.

Pre-flight checklist for work sessions — loads context, runs health scripts, surfaces warnings, and gates on acknowledgment before work begins.

PR Review Retrospective -- actionable analysis of a PR's review cycle with convergence-loop deliverable verification, interactive findings walkthrough, and dashboard for missing retros

Process external PR code review feedback (CodeRabbit, Qodo, SonarCloud, Gemini) through a structured 8-step protocol. Parses all items, categorizes by severity and origin, fixes in priority order, tracks deferred items to TDMS, and captures learnings. This skill processes review feedback — it does not generate reviews.

Query the Content Analysis System (CAS) knowledge base. Search extraction candidates, filter by tags/type/source, search across all analyzed sources.

Content analysis for documents — PDFs, gists, articles, markdown, arxiv papers, code snippets, meeting notes. Dual-lens (Creator View + Engineer View). Two user-invokable depths (Standard / Deep); Quick Scan is triage- only. Outputs to .research/analysis/<doc-slug>/.

Content analysis for video and audio — YouTube, TikTok, podcasts, audio files. Transcription-first pipeline (captions API, user transcript, or Whisper opt-in). Dual-lens (Creator View + Engineer View). Two user- invokable depths (Standard / Deep); Quick Scan is triage-only. Outputs to .research/analysis/<media-slug>/.

Dual-lens repo analysis: Creator View (knowledge, insights, home-repo comparison) + Engineer View (health, security, process). Two user-invokable depths (Standard / Deep); Quick Scan is triage-only. Link mining for curated lists. Fit separation via dual scoring lenses. Outputs to .research/analysis/<repo-slug>/.

Session closure pipeline — context preservation, compliance review, metrics capture, and final commit. Ensures the next session starts with accurate context and current metrics.

Creator-first website analysis skill. Dual-lens (Creator View + Engineer View) three-tier (Quick/Standard/Deep) analysis of arbitrary websites with superpowers-chrome extraction, compliance gates, and multi-mode operation (Page/Site/Expedition/Cross-site).

Cross-source synthesis across all analyzed sources (repos, websites, documents, media). Produces emergent themes, ecosystem gaps, reading chains, mental model evolution, fit portfolio, knowledge map, and an interactive opportunity matrix that routes to the next action. Replaces /repo-synthesis, /website-synthesis, and the cross-type synthesis stub.

Content Analysis System router. Feed it a URL or file path — repo, website, YouTube link, PDF, or gist. Auto-detects type, dispatches to the right handler, and indexes results for /recall.

Create distinctive, production-grade frontend interfaces with high design quality. Use this skill when the user asks to build web components, pages, or applications. Generates creative, polished code that avoids generic AI aesthetics.

DEPRECATED — use /synthesize instead. /repo-synthesis was consolidated into the unified /synthesize skill which handles all 4 source types (repo, website, document, media). Redirect expires next session.

SoNash project context injected into agent definitions via skills: field

DEPRECATED — use /synthesize instead. /website-synthesis was consolidated into the unified /synthesize skill which handles all 4 source types (repo, website, document, media). Redirect expires next session.

Interactive behavioral quality audit for individual skills. Evaluates against 12 quality categories (including T25 convergence loop and completion verification) to surface attention management issues and produce actionable decisions. Includes self-audit verification that all decisions were actually implemented. Uses convergence loops in its own discovery phase. Produces a decision record and updated skill files.

Structured workflow for creating or updating skills — from discovery through validation and audit. Produces structurally correct, behaviorally effective skill packages (SKILL.md + companions) using deep-plan-style discovery with skill-specific question categories and convergence-loop verification of codebase claims.

Lightweight health signal with scoring, benchmarks, trends, and interactive alert-by-alert workflow. Default mode (--limited) checks 18 categories. Use --full for comprehensive reporting with all 42 categories.

Hybrid agent audit — automated structural checks + interactive behavioral review with audit-team (reviewer + fixer) for all agent definitions in .claude/agents/. Convergence-loop verified at every stage.

Creative discovery phase for exploring ideas, directions, and design space before implementation planning or research. Uses Socratic dialogue, research-grade agents, and contrarian challenge to move from seed ideas to chosen directions with structured BRAINSTORM.md output that feeds directly into /deep-plan or /deep-research.

Interactive wizard for creating any type of audit — code, process, infrastructure, ecosystem, or custom. Uses extensive multi-agent discovery and deep-plan Q&A to produce audit skills, multi-AI templates, and TDMS-ready output through a structured, approval-gated workflow.

Comprehensive audit of data system effectiveness — lifecycle scoring across Capture, Storage, Recall, and Action dimensions for all data systems in the project.

Comprehensive diagnostic of the documentation ecosystem — 16 categories across 5 domains with composite health scoring, trend tracking, patch suggestions, and interactive finding-by-finding walkthrough. Covers index sync, link integrity, content quality, generation pipelines, and documentation coverage.

Interactive health dashboard with 8-category composite scoring, 13-dimension drill-down, trend tracking, and fix-on-the-spot triage. Runs 10 health checkers, persists scores to JSONL, and guides interactive remediation through deep-plan Q&A style decisions.

Comprehensive diagnostic of the health monitoring ecosystem — 26 categories across 6 domains with composite health scoring, trend tracking, live test execution, patch suggestions, and interactive finding-by-finding walkthrough.

Comprehensive diagnostic of the hook ecosystem — 20 categories across 6 domains with composite health scoring, trend tracking, patch suggestions, and interactive finding-by-finding walkthrough. Covers Claude Code hooks, pre-commit pipeline, AND CI/CD pipeline health.

Comprehensive diagnostic of the PR review ecosystem — 18 categories across 5 domains with composite health scoring, trend tracking, patch suggestions, and interactive finding-by-finding walkthrough. Triggers: "pr ecosystem audit", "audit pr ecosystem", "review health", "ecosystem health".

Comprehensive diagnostic of the script infrastructure — 18 categories across 5 domains with composite health scoring, trend tracking, patch suggestions, and interactive finding-by-finding walkthrough. Covers all .js files under scripts/ and their npm registrations for module consistency, safety patterns, reachability, code quality, and testing.

Comprehensive diagnostic of the Session Ecosystem — 17 categories across 5 domains with composite health scoring, trend tracking, patch suggestions, and interactive finding-by-finding walkthrough. Covers session lifecycle hooks, state persistence, compaction resilience layers, cross-session safety, and integration configuration.

Comprehensive diagnostic of the skill ecosystem — 21 categories across 5 domains with composite health scoring, trend tracking, patch suggestions, and interactive finding-by-finding walkthrough. Covers all skills, cross-references, bloat detection, and agent orchestration health.

Comprehensive diagnostic of the Technical Debt Management System — 16 categories across 5 domains with composite health scoring, trend tracking, patch suggestions, and interactive finding-by-finding walkthrough. Covers all 37 TDMS scripts, data files, pipeline integrity, and reporting accuracy.

Cross-session todo management with interactive menu, JSONL-backed persistence, and AI-powered context capture. Manages ideas, tasks, and insights that emerge during sessions so they survive across sessions. Independent of GSD.

GitHub platform health assessment with 7 phases, per-finding triage (Fix/Defer/Skip/Suppress), scoped single-phase mode, session-begin quick check, and inline fix execution. Covers security alerts, CI/CD, dependencies, config, releases, insights, and PR health.

Run a comprehensive multi-stage automation audit with parallel agents

Scan all docs, auto-fix formatting/headers/links, report issues as JSONL, and generate enhancement recommendations

Run all 9 domain audits in staged waves and aggregate results

Run a comprehensive enhancement audit across the entire project - code, product, UX, content, workflows, infrastructure, external services, and meta-tooling.

Multi-pass verification of claims about reality using agent-based discovery, T20 tallies, and composable behaviors. Use when claims could waste effort if wrong — audits, discovery, doc-sync, plan verification.