dtsong

Use to convert a Word .docx file to PDF and/or verify its page count. Triggers on: converting docx to pdf, rendering a document, checking how many pages a docx produces, or asserting a page-count constraint (e.g. a resume must stay 2 pages). Wraps LibreOffice headless conversion.

Use whenever Alan Yu is applying to a specific job and needs his resume or cover letter tailored to that opportunity. Triggers on: pasting a job description, asking 'how should I tailor this for [role]', requesting a cover letter for a specific company, asking for fit analysis between Alan's background and a posted role, or producing a .docx resume targeted at a specific employer. Do NOT use for unrelated resume work or for other people's resumes.

Use this skill when scoring or comparing a generated diagram against a human reference. Triggers on "score this diagram", "evaluate my diagram", "compare to reference", or "how accurate is this". Applies when both a generated diagram and a reference image exist and quality assessment is needed. Do NOT use for creating new diagrams (use generate-diagram) or plotting data (use generate-plot).

Use this skill when creating a methodology diagram from research text. Triggers on "make a diagram", "visualize this methodology", "diagram this process", or "generate a figure from this paper". Applies to methodology descriptions, process flows, and research paper sections. Do NOT use for scoring existing diagrams (use evaluate-diagram) or plotting data from CSV/JSON (use generate-plot).

Use this skill when creating a statistical plot or chart from a data file. Triggers on "plot this data", "make a chart", "graph this CSV", or "visualize these results". Applies to CSV, JSON, or tabular data needing bar charts, scatter plots, line graphs, or similar visualizations. Do NOT use for methodology diagrams from text (use generate-diagram) or diagram scoring (use evaluate-diagram).

Use when planning or reviewing Verilator-based simulation workflows for SystemVerilog designs. Covers lint analysis, simulation setup, trace/waveform debugging, coverage-driven verification, and C++ co-simulation. Do not use for commercial EDA tools (use verification-methodology) or RTL design flow (use chip-design-flow).

Use when auditing accessibility compliance of a feature or codebase. Covers WCAG 2.2 AA conformance, screen reader compatibility, keyboard navigation, focus management, color contrast, reduced motion, and ARIA usage. Do not use for general UX review (use journey-mapping) or component interaction specs (use interaction-design).

Use when designing distributed systems or evaluating distributed architecture patterns. Covers CAP theorem trade-offs, consensus protocols (Raft, Paxos), saga orchestration, CRDTs, event sourcing, partition handling, distributed transactions, and failure detectors. Do not use for general API design (use api-design) or database schema design (use schema-design).

Local git operations for syncing, branching, merging, and conflict resolution

GitHub interactions for issues, PRs, releases, and repository management

Use this skill when performing hardware security analysis for System-on-Chip components — threat modeling, verification scaffolding, compliance mapping, executive briefing, microarchitectural attack analysis, physical side-channel assessment, kernel security analysis, emerging hardware security, or TLA+ formal specification. Routes to the appropriate specialist. Trigger phrases include "threat model my SoC", "run STRIDE analysis", "generate SVA assertions", "compliance check against FIPS", "executive summary of findings", "Spectre analysis for cache", "DPA attack assessment", "kernel hardening review", "PQC hardware review", "TLA+ spec for access control". Do NOT use for software-only security, network security, or web application security.

Use when working with Terraform or OpenTofu - creating modules, writing tests (native test framework, Terratest), setting up CI/CD pipelines, reviewing configurations, choosing between testing approaches, debugging state issues, implementing security scanning (trivy, checkov), or making infrastructure-as-code architecture decisions

Use this skill when scoring or comparing a generated diagram against a human reference. Triggers on "score this diagram", "evaluate my diagram", "compare to reference", or "how accurate is this". Applies when both a generated diagram and a reference image exist and quality assessment is needed. Do NOT use for creating new diagrams (use generate-diagram) or plotting data (use generate-plot).

Use this skill when creating a methodology diagram from research text. Triggers on "make a diagram", "visualize this methodology", "diagram this process", or "generate a figure from this paper". Applies to methodology descriptions, process flows, and research paper sections. Do NOT use for scoring existing diagrams (use evaluate-diagram) or plotting data from CSV/JSON (use generate-plot).

Use this skill when creating a statistical plot or chart from a data file. Triggers on "plot this data", "make a chart", "graph this CSV", or "visualize these results". Applies to CSV, JSON, or tabular data needing bar charts, scatter plots, line graphs, or similar visualizations. Do NOT use for methodology diagrams from text (use generate-diagram) or diagram scoring (use evaluate-diagram).

Use this skill when scoping a new consulting engagement — "help me scope this project," "I need an engagement brief," or gathering client context through conversational clarification. Transforms natural-language descriptions into structured engagement briefs with targeted questions, sensitivity inference, and stakeholder mapping. Do NOT use for technology research (research-skill), risk assessment (assessment-skill), or client-facing deliverables (deliverable-skill).

Security-first CI/CD workflow generation (GitHub Actions).

Build minimal, secure Dockerfiles with verification-first workflow.

Fast git snapshot for status, diff, and recent history.

Safe day-to-day git workflows for branching, sync, and cleanup.

Practical GitHub issue and pull request workflows using gh CLI.

Generate minimal-diff Helm values with schema-aware validation.

Project convention references for Python, TypeScript, and Terraform.

Terraform/OpenTofu authoring, testing, and security workflow.

Use when a user reports a frontend bug, visual defect, or unexpected behavior in a Next.js/TypeScript application. Orchestrates a multi-phase QA pipeline — component mapping, diagnosis, fix, and regression testing — by classifying symptoms and dispatching to the appropriate specialist skill. Does not perform diagnosis or fixes directly; delegates all domain work to specialists. Not for backend-only issues, infrastructure problems, or build/deployment failures.

Use when investigating any frontend issue before diagnosis or debugging — maps a Next.js route to its full component tree. Resolves "what components render on this page," "show me the component tree for /dashboard," or "trace imports for this route." Produces a ComponentMap artifact with server/client boundaries, props, hooks, data fetching, and styling metadata consumed by all downstream QA skills. Not for diagnosing bugs, fixing code, or generating tests — those are handled by specialist skills after mapping is complete.

TRIGGER_DESCRIPTION. Use when USER_CONTEXT. Routes to specialists for CAPABILITIES.

TRIGGER_DESCRIPTION. Use when USER_CONTEXT. Covers CAPABILITIES.

Use when a diagnosed fix needs to be applied and verified against a Next.js component — "apply this fix," "fix and run tests," or "verify this change." Applies a single diagnosed fix with mandatory diff preview, then runs scoped and broad verification (tsc, lint, tests) with atomic safety and revert capability. Not for diagnosing bugs (use ui-bug-investigator or css-layout-debugger) or generating regression tests (use regression-test-generator).

Use when a user reports CSS, layout, or styling issues in a Next.js application — "element overlapping," "spacing is wrong," "responsive layout broken," "dark mode not working," "Tailwind classes not applying," or "alignment off." Diagnoses through a 6-phase pipeline that mirrors browser evaluation order, covering token resolution, cascade conflicts, layout models, stacking contexts, and viewport responsiveness. Handles Tailwind, CSS Modules, styled-components, and vanilla CSS. Not for JavaScript logic bugs, data fetching issues, or server-side rendering errors — route those to ui-bug-investigator.

Use when a verified fix needs a regression test — "write a test for this fix," "generate regression test," or "add test coverage for this bug." Auto-detects project test conventions, routes to Vitest+RTL or Playwright based on component type (sync vs async server), generates targeted tests with anti-brittleness constraints, and annotates coverage boundaries. Not for diagnosing bugs, applying fixes, or writing feature tests — use ui-bug-investigator, component-fix-and-verify, or manual test authoring respectively.

Use when a user reports a non-CSS UI bug in a Next.js App Router application — "blank page," "data not loading," "click does nothing," "hydration mismatch," "flicker on navigation," or "state not updating." Runs symptom-targeted diagnostic checks against a component map to identify rendering, state, event handling, and data flow issues. Not for CSS layout, styling, or visual design problems — route those to css-layout-debugger.

Use when working with dbt (data build tool) - creating models, writing tests, CI/CD pipelines, materializations, sources, staging/intermediate/marts layers, Snowflake/BigQuery warehouse configuration, incremental strategies, Jinja macros, data quality, semantic layer, or making analytics engineering decisions

Local git operations for syncing, branching, merging, and conflict resolution

GitHub interactions for issues, PRs, releases, and repository management

Test-driven development reference for writing good tests, designing testable interfaces, mocking at system boundaries, and refactoring after green. Use when writing tests, reviewing test quality, or applying red-green-refactor workflow. Not for running test suites or CI configuration — use language-conventions or cicd-generation for those.

Use when working with Terraform or OpenTofu - creating modules, writing tests (native test framework, Terratest), setting up CI/CD pipelines, reviewing configurations, choosing between testing approaches, debugging state issues, implementing security scanning (trivy, checkov), or making infrastructure-as-code architecture decisions

Security audit checklist for web applications. Use when reviewing, auditing, or hardening a web app's security posture. Covers rate limiting, auth headers, IP blocking, CORS, security middleware, input validation, file upload limits, ORM usage, and password hashing. Triggers on requests like "review security", "harden this app", "security audit", "check for vulnerabilities", or when building/reviewing API endpoints.

Use this skill when performing hardware security analysis for System-on-Chip components — threat modeling, verification scaffolding, compliance mapping, executive briefing, microarchitectural attack analysis, physical side-channel assessment, kernel security analysis, emerging hardware security, or TLA+ formal specification. Routes to the appropriate specialist. Trigger phrases include "threat model my SoC", "run STRIDE analysis", "generate SVA assertions", "compliance check against FIPS", "executive summary of findings", "Spectre analysis for cache", "DPA attack assessment", "kernel hardening review", "PQC hardware review", "TLA+ spec for access control". Do NOT use for software-only security, network security, or web application security.

Performs bank reconciliations between a bank statement export and GL cash detail. Use when the user provides a bank statement file (CSV, XLSX, or PDF) and a GL cash detail file and asks to reconcile, match transactions, identify outstanding items, or produce a bank rec workpaper. Outputs a matched/unmatched workbook, aged reconciling items, proposed adjusting JEs, and an evidence table.

Loads firm-specific accounting configuration including chart of accounts, materiality thresholds, close calendar, approval matrix, and standard mappings. Use at the start of any workflow that needs to apply firm-specific rules rather than generic accounting conventions.

Performs month-end flux/variance analysis against prior period and budget. Use when the user provides a trial balance and asks to analyze variances, explain fluxes, produce variance commentary, or build a flux workpaper. Applies firm materiality thresholds, calculates dollar and percent changes, categorizes variance type, drafts explanations where support exists, and generates targeted preparer questions where it doesn't.

Reviews proposed journal entries for posting readiness. Use when the user provides a list or file of journal entries and asks to review, validate, check for errors, or approve for posting. Verifies debit/credit balance, account coding, support adequacy, memo quality, period appropriateness, duplicate risk, and unusual patterns. Classifies each entry as ready, needs clarification, or reject.

Maps audit PBC request lists to support files, identifies gaps, drafts auditor responses, and assembles audit-ready evidence packages. Use when the user provides a PBC request list and a folder of support files and asks to match support to requests, identify missing items, flag duplicates or orphaned files, or draft responses to the auditor.

Builds and maintains prepaid expense amortization schedules. Use when the user provides vendor invoices, service period information, or an existing prepaid template and asks to extract prepaid items, build amortization schedules, calculate monthly expense, or generate journal entries for prepaid recognition and amortization. Produces a formula-driven schedule, not hardcoded tables.

Use whenever producing an accounting workpaper, schedule, reconciliation, journal entry, or audit deliverable. Defines firm formatting conventions, the evidence-table requirement, journal entry templates, and the standard for source tie-outs. Required dependency for bank-rec, flux-analysis, prepaid-schedule, pbc-package, and je-review skills.

Use when orchestrating monthly or quarterly close. Covers the close calendar, sub-ledger cutoffs, accruals, reclass entries, and post-close review. Do not use for individual reconciliations (use finance-reconciliation) or journal-entry construction (use finance-journal-entries).

Use when testing a SOX 404 control or designing a controls walkthrough. Covers control description, design effectiveness, operating effectiveness, deficiency evaluation, and remediation. Do not use for evidence-package preparation (use finance-evidence-package — Phase 3+).

Use when constructing a GAAP/IFRS journal entry from facts. Covers debit/credit selection, account coding, sub-ledger impact, and supporting memo. Do not use for reconciliation (use finance-reconciliation) or close orchestration (use finance-close-checklist).

Use when reconciling a balance-sheet account during the close cycle. Covers sub-ledger to GL ties, intercompany eliminations, exception triage, and audit-ready documentation. Do not use for journal-entry construction (use finance-journal-entries) or close orchestration (use finance-close-checklist).

Use when answering a federal, state, or international tax question that requires authoritative research. Covers IRC, Treasury Regs, Rev Rul/Proc, case law, and applicable treaty positions. Do not use for tax-provision computation (use finance-provision — Phase 3+).

Use when explaining actual-vs-plan or actual-vs-prior variance with commentary. Covers volume/price/mix decomposition, run-rate impact, and management commentary. Do not use for forecast revision (use finance-forecast-update — Phase 3+) or close orchestration (use finance-close-checklist).

Use when the council needs to surface organizational knowledge buried across multiple internal sources (wikis, design docs, ADRs, past tickets, postmortems, chat archives, code repos). Plans where to look, what to cross-reference, and how to synthesize findings into evidence the council can act on. Do not use for external market research (use competitive-analysis), library evaluation (use library-evaluation), or technology trend assessment (use technology-radar).

Use when a user asks to read, analyze, summarize, or extract from a heavyweight file (PDF, DOCX, PPTX, XLSX, CSV, TSV). Converts the file into markdown or CSV first, generates a lightweight index, and only spends model tokens on the compressed artifact. Trigger on "read this PDF", "look through this spreadsheet", "summarize this deck", or any time raw file ingestion would waste tokens. Do NOT trigger for plain text, markdown, JSON, YAML, or source code files — those are cheap to read directly.

TRIGGER_DESCRIPTION. Use when USER_CONTEXT. Routes to specialists for CAPABILITIES.

TRIGGER_DESCRIPTION. Use when USER_CONTEXT. Covers CAPABILITIES.

SPECIALIST_TRIGGER. Use when SPECIALIST_CONTEXT.