diShine

Account health monitor — NPS signal, renewal risk, expansion opportunity, comms cadence.

Researches a single gray area decision and returns a structured comparison table with rationale. Spawned by discuss-phase advisor mode.

Deeply analyzes codebase for a phase and returns structured assumptions with evidence. Spawned by discuss-phase assumptions mode.

Audits a draft against a client's documented brand voice and tone axes — produces red/amber/green rating + specific rewrites.

Generates interactive charts and diagrams via ECharts MCP, Mermaid MCP, or VegaLite MCP. Produces publication-ready visualizations from data or specifications.

Prospect/client intelligence — company dossier from verified live sources only (Apollo, Hunter, LinkedIn via Playwright, company site via WebFetch).

Explores codebase and writes structured analysis documents. Spawned by map-codebase with a focus area (tech, arch, quality, concerns). Writes documents directly to reduce orchestrator context load.

Competitive intelligence — maps competitor positioning, pricing, differentiation, weaknesses from public sources.

Structural + line editor — tightens prose, enforces voice, flags weak claims, preserves meaning.

Long-form copywriter — landing pages, email sequences, ads — conversion-focused, voice-faithful.

CRM hygiene + enrichment — dedupes, fills gaps, syncs stages, purges stale records. Respects GDPR.

Interprets datasets — cleans, hypothesizes, tests, visualizes; grounds every insight in the numbers.

Local data analysis via DuckDB/SQLite MCP — SQL on CSV/Parquet/JSON/Excel without cloud services. Produces charts via ECharts/Mermaid MCP when available.

Investigates bugs using scientific method, manages debug sessions, handles checkpoints. Spawned by /shine-debug orchestrator.

Verifies factual claims in generated docs against the live codebase. Returns structured JSON per doc.

Writes and updates project documentation. Spawned with a doc_assignment block specifying doc type, mode (create/update/supplement), and project context.

Executes SHINE plans with atomic commits, deviation handling, checkpoint protocols, and state management. Spawned by execute-phase orchestrator or execute-plan command.

GDPR/privacy analyst — cookie audit, PII classification, DPIA scaffold, legal-basis decision tree. Never gives legal advice; flags for counsel.

Infrastructure operations via Docker MCP, Kubernetes MCP, and monitoring MCPs (Signoz, VictoriaMetrics). Manages containers, clusters, and observability.

Verifies cross-phase integration and E2E flows. Checks that phases connect properly and user workflows complete end-to-end.

Scores a lead list against an ICP — firmographics + behavioural + intent signals; never fabricates data.

Designs a MarTech stack — collection, activation, analytics, automation — vendor-neutral, GDPR-native.

Fills Nyquist validation gaps by generating tests and verifying coverage for phase requirements

Builds buyer personas from qualitative research — interviews, surveys, public profiles.

Researches how to implement a phase before planning. Produces RESEARCH.md consumed by shine-planner. Spawned by /shine-plan-phase orchestrator.

Verifies plans will achieve phase goal before execution. Goal-backward analysis of plan quality. Spawned by /shine-plan-phase orchestrator.

Creates executable phase plans with task breakdown, dependency analysis, and goal-backward verification. Spawned by /shine-plan-phase orchestrator.

Cross-client PM — tracks deadlines, dependencies, blockers; produces daily/weekly stand-up inputs.

Researches domain ecosystem before roadmap creation. Produces files in .planning/research/ consumed during roadmap creation. Spawned by /shine-new-project or /shine-new-milestone orchestrators.

Long-form proposal author — MoSCoW + MD estimates + ~15% discount, Italian by default, RAG-grounded, never ships without a proposed call.

Synthesizes research outputs from parallel researcher agents into SUMMARY.md. Spawned by /shine-new-project after 4 researcher agents complete.

Runs structured retrospectives — Start/Stop/Continue, 5-whys on friction, action items with owners.

Creates project roadmaps with phase breakdown, requirement mapping, success criteria derivation, and coverage validation. Spawned by /shine-new-project orchestrator.

Account planning — maps buying committee, entry points, multi-thread sequence, expansion path.

Executes untrusted or experimental code in isolated containers via Docker MCP or Microsandbox. Never runs risky code directly on the host.

Verifies threat mitigations from PLAN.md threat model exist in implemented code. Produces SECURITY.md. Spawned by /shine-secure-phase.

SEO strategy — technical audit, keyword gap vs competitors, prioritized fix list. Uses Ahrefs MCP, Lighthouse CLI, GSC if available.

Multilingual translator (IT / EN / FR / ES) — preserves voice, terminology, CTA impact, cultural nuance.

Retroactive 6-pillar visual audit of implemented frontend code. Produces scored UI-REVIEW.md. Spawned by /shine-ui-review orchestrator.

Validates UI-SPEC.md design contracts against 6 quality dimensions. Produces BLOCK/FLAG/PASS verdicts. Spawned by /shine-ui-phase orchestrator.

Produces UI-SPEC.md design contract for frontend phases. Reads upstream artifacts, detects design system state, asks only unanswered questions. Spawned by /shine-ui-phase orchestrator.

Analyzes extracted session messages across 8 behavioral dimensions to produce a scored developer profile with confidence levels and evidence. Spawned by profile orchestration workflows.

Verifies phase goal achievement through goal-backward analysis. Checks codebase delivers what phase promised, not just that tasks completed. Creates VERIFICATION.md report.

Scans code and dependencies for security vulnerabilities using Semgrep MCP (SAST) + OSV MCP (dependency audit) + sslmon MCP (SSL/domain). Produces actionable security report.

Deep web research with tiered search: SearXNG (free) → Brave/Tavily (freemium, ask) → Perplexity/Exa (paid, explicit approval). Grounds every claim in a retrieved, cited source.

Design an A/B test — hypothesis, metric, MDE, sample size, duration, guardrails.

Add an idea to the backlog parking lot (999.x numbering)

Add phase to end of current milestone in roadmap

Generate tests for a completed phase based on UAT criteria and implementation

Capture idea or task as todo from current conversation context

Analyze phase dependencies and suggest Depends on entries for ROADMAP.md

Design a REST/GraphQL API surface — resources, operations, errors, pagination, auth.

Sketch an architecture diagram in Mermaid or draw.io from a description.

Choose and configure an attribution model — goals, channels, tooling, limitations.

Audit milestone completion against original intent before archiving

Cross-phase audit of all outstanding UAT and verification items

Run all remaining phases autonomously — discuss→plan→execute per phase

Draft a long-form blog post (900–1500 words) with SEO front-matter, H-tree, FAQ, and JSON-LD.

Define or audit brand voice — tone axes, do/don't examples, voice against samples.

Write a campaign brief — objective, audience, message, channels, KPIs, budget, timeline.

Project-team capacity for the next N weeks — commitments vs availability, conflict flags.

Assemble a client case study — challenge, approach, solution, results, quote, visual brief.

Draft a changelog entry from recent commits — Keep-a-Changelog format.

Draft a change management plan — Kotter 8-step or ADKAR, tailored to the change scope.

Generate interactive charts (ECharts), diagrams (Mermaid), or statistical visualizations (VegaLite) from data or specifications.

List pending todos and select one to work on

Archive accumulated phase directories from completed milestones

Produce a 1-page briefing on a client before a meeting — pulls context from memory + last emails + Asana.

Review a draft against the client's language / tone / CC conventions before sending.

Draft a compliant cold email — personalized hook, value line, soft CTA, P.S. — never auto-send.

Produce a competitor map — positioning, pricing, differentiation, weakness, how we win.

Archive completed milestone and prepare for next version

Generate AI-use policy + DPIA skeleton for a client, based on ai-compliance-framework templates.

Build a monthly content calendar — themes, channels, owners, publish dates, CTA per post.

Scan a URL for cookies and classify them (strictly-necessary / preferences / statistics / marketing).

Specify a BI dashboard — audience, KPIs, filters, cadence, data sources.

Draft a data contract between producer and consumer — schema, SLA, breaking changes.

Run SQL analytics on local data files (CSV, Parquet, JSON, Excel) via DuckDB or SQLite MCP. No cloud — all local.

Systematic debugging with persistent state across context resets

Audit project dependencies for known CVEs using OSV MCP or native package auditors (npm audit, pip audit, govulncheck).

Pre-deploy checklist — tests, migrations, feature flags, rollback, monitoring, comms.

Structured design critique — hierarchy, legibility, consistency, accessibility, brand fit.

Run the diShine digital maturity scorecard — 5 dimensions, personalized roadmap.

Run a 45-min discovery call structure — context, pain, impact, capacity, next step.

Post-discovery synthesis — context, pains, options, recommended path, MD estimate.

Gather phase context through adaptive questioning before planning. Use --auto to skip interactive questions (Claude picks recommended defaults). Use --chain for interactive discuss followed by automatic plan+execute. Use --power for bulk question generation into a file-based UI (answer at your own pace).

Generate or update project documentation verified against the codebase

Route freeform text to the right SHINE command automatically

Draft a client-facing email in Kevin's style — warm opening, bullets, CLIENT | Topic subject, internal CC logic.

Compress a long document into a 1-page executive summary — decision-ready.

Execute all plans in a phase with wave-based parallelization

Execute a trivial task inline — no subagents, no planning overhead

Craft the right follow-up on a dormant thread — context-aware, not annoying, with a clear out.

Post-mortem investigation for failed SHINE workflows — analyzes git history, artifacts, and state to diagnose what went wrong

Audit a GA4 property — events, conversions, data retention, PII, consent, filters.

Route a GDPR / consent / PII task through the right diShine tool BEFORE drafting. Fires CLAUDE.md Rule #18.

Diagnose planning directory health and optionally repair issues

Show available SHINE commands and usage guide

Define an Ideal Customer Profile — firmographics, triggers, no-go list, scoring rubric.

Write a blameless post-mortem — timeline, impact, root cause, action items.

Insert urgent work as decimal phase (e.g., 72.1) between existing phases

Draft a freelance / agency invoice — line items from MD log, VAT, payment terms.

Join the SHINE Discord community

Initialize a new client engagement — memory file, Asana project, kickoff agenda, welcome email draft.

Build a KPI tree — north star, driver metrics, input metrics, ownership.

Write a conversion-focused landing page — hero, value prop, social proof, objections, CTA.

Run the @lead-enrichment pattern — local scripts first, Apollo/Hunter/Apify supplement, inferred emails labeled.

Capture a learning into a reusable memory entry — preference, style, or external reference.

Draft a LinkedIn connection + follow-up DM sequence — human, short, non-salesy.

Surface Claude's assumptions about a phase approach before planning

List all SHINE workspaces/workstreams with status and active marker.

Write a designer brief for a logo — concept, references, constraints, deliverables.

Interactive command center for managing multiple phases from one terminal

Analyze codebase with parallel mapper agents to produce .planning/codebase/ documents

Structure meeting notes — attendees, decisions, action items, open questions.

Audit page <meta> tags, OpenGraph, and Twitter cards for a URL.

Plan a technical migration — inventory, target, strategy, phases, cutover, rollback.

Generate a comprehensive project summary from milestone artifacts for team onboarding and review

Generate product / feature / company names — brief, 20 candidates, shortlist rationale.

Triage an NDA — flag one-sided clauses, term length, jurisdiction, scope issues.

Network diagnostics using Globalping MCP — ping, traceroute, DNS lookup from global probes. Also checks SSL certs via sslmon MCP.

Start a new milestone cycle — update PROJECT.md and route to requirements

Initialize a new project with deep context gathering and PROJECT.md

Draft a weekly/bi-weekly newsletter issue — from brief to ready-to-paste HTML/MJML.

Create an isolated workspace with repo copies and independent .planning/

Automatically advance to the next logical step in the SHINE workflow

Zero-friction idea capture. Append, list, or promote notes to todos.

Draft quarterly OKRs — 3 objectives × 3 key results, measurable, ambitious, aligned.

Create context handoff when pausing work mid-phase

Build a buyer persona — role, jobs-to-be-done, pains, gains, objections, where they hang out.

Anonymize PII in a dataset before LLM ingestion via dishine-data-safe-usb.

Compose skills into a sequential pipeline — skill-a → skill-b → skill-c

Create phases to close all gaps identified by milestone audit

Create detailed phase plan (PLAN.md) with verification loop

Capture a forward-looking idea with trigger conditions — surfaces automatically at the right milestone

Create a clean PR branch by filtering out .planning/ commits — ready for code review

Draft a press release following AP style — headline, dateline, lead, body, boilerplate, contact.

Draft pricing page copy — tiers, inclusions, feature table, FAQ, call-to-action.

Document a repeatable process — trigger, steps, tools, owner, SLA, exit criteria.

Generate developer behavioral profile and create Claude-discoverable artifacts

Check project progress, show context, and route to next action (execute or plan)

Assemble a client proposal with MoSCoW structure, MD estimates (8h/MD), and a ~15% discount option — Italian by default.

Review a pull request — correctness, security, readability, test coverage, blast radius.

Execute a quick task with SHINE guarantees (atomic commits, state tracking) but skip optional agents

Generate a README following diShine house style — badges, ToC, install, usage, contrib, license.

Reapply local modifications after a SHINE update

Remove a future phase from roadmap and renumber subsequent phases

Remove a SHINE workspace and clean up worktrees

Research how to implement a phase (standalone - usually use /shine-plan-phase instead)

Resume work from previous session with full context restoration

Cross-session retrospective — review learning-log.jsonl and propose memory updates

Run a structured retro on a delivered project — what worked / what didn't / what to systemize into a memory entry.

Review and promote backlog items to active milestone

Request cross-AI peer review of phase plans from external AI CLIs

Draft a response to an RFP — requirements matrix, tailored narrative, pricing, timeline.

Build a risk register — probability × impact, mitigations, owners, review cadence.

Sketch a product or engagement roadmap — Now / Next / Later, with assumptions.

Estimate ROI of an automation / process change — baseline hours, saved hours, break-even, payback period.

Turn a sales call into structured notes — BANT, next steps, objections, CRM-ready.

Prepare for a sales call — attendee research, agenda, killer questions, likely objections.

Structure a sales deck — problem, solution, differentiation, proof, pricing, close.

Execute code safely in an isolated Docker container. Prevents untrusted code from running directly on the host.

Retroactively verify threat mitigations for a completed phase

Scan codebase for vulnerabilities using Semgrep MCP (SAST) + OSV MCP (dependency audit). Free, local, no cloud.

Run a lightweight SEO audit on a URL — routes to meta-inspector + lighthouse-dashboard + SERP check.

Generate a report of what this Claude Code session accomplished (commits, files, phases, tools).

Switch SHINE model profile (quality|balanced|budget|inherit) with validation.

Configure SHINE workflow toggles and model profile

Create PR, run review, and prepare for merge after verification passes

Produce platform-native social posts — LinkedIn / IG / X / Threads variants from one angle.

Map stakeholders — interest × influence, with engagement strategy per quadrant.

Show SHINE project stats: phase counts, commit velocity, workstream status.

Produce a weekly client-facing status report — done / in progress / next / risks / asks.

Facilitate a SWOT — strengths, weaknesses, opportunities, threats — with prioritization.

Re-run SHINE SessionStart hooks (memory symlink, plugin sync, update check).

Analyze a GTM container export via tag-auditor — 14 governance/consent/security checks.

Write a technical spec — problem, goals, non-goals, design, alternatives, rollout, risks.

Design a test strategy — pyramid, coverage targets, fixtures, CI gates.

Manage persistent context threads for cross-session work

Aggregate a week of work logs into a timesheet — by client, by phase, billable %.

Interactive walkthrough of SHINE — verifies install, teaches core concepts, troubleshoots issues

Translate content between IT / EN / FR / ES — preserving voice, terminology, CTAs.

Generate UI design contract (UI-SPEC.md) for frontend phases

Retroactive 6-pillar visual audit of implemented frontend code

Update SHINE to latest version with changelog display

Retroactively audit and fill Nyquist validation gaps for a completed phase

Crystallize a value proposition — job-to-be-done, pains, gains, unique mechanism, proof.

Store and retrieve semantic memories using Qdrant MCP vector database. Enables cross-session knowledge recall.

Vendor due-diligence checklist — security, privacy, SLA, pricing, exit clause.

Validate built features through conversational UAT

Plan a webinar — topic, audience, agenda, promotion plan, follow-up sequence.

Deep web research using tiered search — SearXNG (free) → Brave/Tavily (ask) → Perplexity (approve). Every claim grounded in a cited, fetched source.

Plan the week from open commitments — priorities, time blocks, risks.

Manage parallel workstreams — list, create, switch, status, progress, complete, and resume