CSalcedoDataBI
UserThe canonical DAX language reference for AI agents — every function with its signature, semantics, and the gotchas the documentation doesn't tell you.
Categories
Indexed Skills (18)
dax-lib-install
Use when a DAX UDF already published on daxlib.org (found via the dax-lib index) needs to be actually installed against a real model — not just documented. Fetches the real function body from the daxlib/daxlib registry, checks its declared license, installs it (modeling MCP if connected, else a direct TMDL edit), runs a real DAX query to confirm it executes, and leaves it attributed on the function itself (author, license, source URL) so it is never mistaken for original work. Triggers on "install this UDF", "bring in the daxlib function", "add TimeSeries.MovingAverage to my model", "vendor this DAX package".
dax-lib
Use when looking for an EXISTING published DAX user-defined function before writing one — moving averages, format strings, SVG visuals, KPI/variance, RFM or ABC classification, unit conversion, model audit/metadata, sample data, time intelligence. Offline index of the DAX Lib registry (daxlib.org) — what exists, who wrote it, and where to get it. Triggers on "is there a UDF for", "find an existing DAX function", "ready-made", "already exists", "daxlib", "DAX package".
dax-reference
Use when you need to know what a built-in DAX function does, its exact signature, what it returns, whether it is legal in a measure, a calculated column, a calculated table or a visual calculation, whether Microsoft discourages it, or which of several similar functions to reach for — and equally for the DAX language itself when no single function is in question, such as evaluation context, context transition, filter and row context, the query statements EVALUATE, DEFINE, MEASURE, ORDER BY and START AT, operators, data types, BLANK semantics, variables with VAR, and Microsoft's own best-practice guidance. The canonical reference for DAX. Triggers on "what does X do in DAX", "DAX function signature", "which DAX function", "difference between", "is X deprecated", "can I use X in a calculated column", "does X work in a measure", "what is evaluation context", "context transition", "DAX query syntax", "EVALUATE statement", "what does BLANK mean in DAX", "DAX operators", "DAX best practice".
dax-udf-authoring
Use when AUTHORING your own DAX user-defined function — the mechanics of writing a FUNCTION — declaring parameter types (Scalar, AnyRef, ColumnRef, MeasureRef, CalendarRef), VAL vs EXPR evaluation, TABLEOF/NAMEOF, optional parameters and defaults, dot-notation naming, GA limitations and parser red-underline bugs. Not for finding an existing one (that is dax-lib) nor for what a built-in does (that is dax-reference). Triggers on "write a DAX UDF", "define a DAX FUNCTION", "ANYREF", "VAL vs EXPR", "TABLEOF", "optional parameter in a UDF".
dax-window-functions
Use when writing a DAX window function — WINDOW, OFFSET, INDEX, RANK, ROWNUMBER, MOVINGAVERAGE, RUNNINGSUM with ORDERBY/PARTITIONBY/MATCHBY — for rolling and trailing averages, running totals, prior-row or prior-year comparisons, or ranking within a partition. Covers ABS vs REL positioning, the ALLSELECTED relation default that silently returns blanks, and MATCHBY for fact tables. Triggers on "rolling average", "running total", "trailing N", "OFFSET returns blank", "ABS vs REL", "PARTITIONBY".
abios-feedback
Use when, while working in ANY project (especially a PRIVATE one), you notice a bug or improvement for the agentic-board tool itself. Captures it as a SANITIZED issue on the tool's OWN public repo/board — never touching the current project and never leaking private data. Triggers — "mejora para la herramienta", "esto es una mejora para agentic-board", "abios bug", "esto deberíamos arreglarlo en el plugin", a guard block, a recurring gh/board failure.
board-expert
Auto-expert engine for /agentic-board:expert — take a tracked plan and execute it autonomously through a specialized expert persona (research, build, test with recorded evidence, self-drive the board), stopping only at the irreversible line. Owns the config (build the contract) and auto (launch the autonomous run) recipes. Routed by the expert command; never typed directly. Triggers — "/expert", "/board expert", "auto-experto", "corre el experto autónomo", "ejecuta el plan solo".
diagram-authoring
Use whenever a diagram goes into a plugin-produced markdown artifact — a /board plan or epic/issue body, a handoff, a status update, KNOWLEDGE.md, the README, or a blog entry. Emit diagram-as-code (Mermaid by default), NEVER hand-drawn ASCII art. Covers flowcharts, architecture, sequence, state, ER and gantt diagrams; escalate to D2/Graphviz (rendered via Kroki) only when a graph is too dense for Mermaid. Triggers — "draw a diagram", "add a flowchart/architecture/sequence/state diagram", "diagrama", "haz un diagrama", "visualiza este flujo", "add a diagram to the plan/issue/handoff/README".
gh-account
Use FIRST before any GitHub Projects/issues operation in the agentic-board suite. Resolves which account (default CSalcedoDataBI, override PAL-Devs) and reads its PAT from the Windows user registry, injecting GH_TOKEN per-invocation without touching `gh auth switch`. Triggers — any board/issue op, "cambia a CSalcedoDataBI", 403 on a PAL board, INSUFFICIENT_SCOPES/read:project.
knowledge-registry
Use to capture and read a project's knowledge references by domain — add a research MD, repo, doc folder, URL, NotebookLM notebook or video to knowledge/registry.json, regenerate the KNOWLEDGE.md table, or publish it to the repo's GitHub Wiki. Distinct from MEMORY.md (agent facts) and HANDOFF.md (task resume). Triggers — "guarda esta referencia", "agrega a knowledge", "registra este link/doc/repo", "muéstrame la tabla de conocimiento", "publica el knowledge al wiki", "/knowledge add", "/knowledge list", "/knowledge gen", "/knowledge wiki".
project-scan
Use to scan the CURRENT project for latent, hard-to-track work — code TODO/FIXME, unchecked checklists and "pending/next steps" in docs, and plan/spec docs not yet tracked — then convert the chosen items into issues + a work plan on THIS project's board. Triggers — "escanea el proyecto", "convierte los pendientes en issues", "harvest backlog", "qué hay sin trackear", "arma el plan de trabajo", /scan.
projects-admin
Use to administer or automate a GitHub Projects (v2) board or its issues — create/configure a project, set Status/Priority/Target fields, add/move/bulk-edit items, link a board to a repo, install CI auto-add, run /board fill to detect and fix gaps, or /board work to see pending issues across boards and start one. Always resolves identity via gh-account (default CSalcedoDataBI). Triggers — "administra el board", "mueve a Done", "crea el project", "add to board", "bulk close", "automatiza el board", "llena el board", "qué hay pendiente", "qué issue trabajo", "empecemos un issue", /board.
skills-audit
Use to evaluate Agent Skills and detect when one is failing — a static health audit (empty/first-person/over-budget descriptions, missing triggers, near-duplicates, misplaced) plus an on-demand runtime trigger-eval (run a realistic prompt with the skill enabled vs disabled, 3x, score false positives/negatives). Classifies failures (under-triggered, over-triggered, wrong output, ignored, obsolete) and, behind a human gate, files a SANITIZED issue to the repo that OWNS the skill — the tool's own board for its skills, the project's board for the project's own skills, local-only for third-party — never leaking the private project you are working in. Triggers — "audita mis skills", "esta skill no dispara/falla", "evalúa el triggering", "skill health check", "por qué no se activó la skill", /skills audit.
skills-bootstrap
Use to install curated best-practice tools by PROFILE WITHOUT duplicating what is already installed — the `quality` profile (skill-creator, writing-skills, skill-improver, second-opinion) and the `bi` profile (Microsoft Fabric / Power BI ecosystem, e.g. microsoft/skills-for-fabric). Detects the gap against the live inventory, recommends only what is missing, and installs each from its source (clean skill-clone preserving LICENSE, or the plugin's own install command) — never re-installing what you already have. NOT for organizing or auditing existing skills — use skills-organize or skills-audit for that. Triggers — "instala las skills de buenas prácticas", "bootstrap skills", "instala el toolkit de BI/Fabric", "qué me falta del perfil bi", "setup skill toolkit", /skills bootstrap, /skills bootstrap bi.
skills-organize
Use to inventory and organize Agent Skills in a repo or monorepo — a read-only catalog + health report (description lint, near-duplicate detection, misplaced/orphaned skills) and a propose-then-confirm reorganize that moves scattered SKILL.md files into the canonical .claude/skills/<project>/<skill>/ layout with git mv (dry-run first, fully reversible) and writes a skills-index.json. Use when skills are scattered across a messy monorepo, when you want a catalog of what is installed, or before auditing skill quality. Triggers — "organiza mis skills", "cataloga las skills", "skills regadas/desordenadas", "skills health", "reorganiza el monorepo", "no mezclar skills", /skills organize.
tmdl-review
Review a TMDL semantic-model diff for breaking schema changes before merging a PBIP change. Compares two versions of the *.tmdl files (a PR, or two git refs) and classifies every change as BREAKING / WARNING / INFO. Use when the user wants to know whether a Power BI / Fabric model change is safe to merge. Triggers — "revisa el diff TMDL", "breaking changes del modelo", "compara el modelo semántico", "¿este cambio rompe el modelo?", "review the TMDL diff", "detect breaking schema changes".
tools-catalog
Use to browse, research and install the project's referenced external tools from one unified catalog that merges the knowledge registry (references) with the installable toolkit presets. Install one tool or all missing installables at once, kind-aware (skill-clone preserves LICENSE; a plugin surfaces its own install command, never cherry-picked). The discoverability surface at the intersection of knowledge-ops and skills-ops. Triggers — "/tools", "navega las herramientas referenciadas", "instala esta herramienta", "instálalas todas", "qué herramientas hay para instalar".
knowledge-harvest
Use to sweep the CURRENT repo for knowledge references that are not yet catalogued — docs/*.md research files and http links in README/docs — and turn the chosen ones into registry entries by domain. The batch-capture companion to knowledge-registry, parallel to project-scan. Triggers — "cosecha las referencias", "escanea el repo por docs/links", "qué conocimiento no está registrado", "harvest knowledge", "/knowledge harvest".
Bio shown is the top-scored skill's repo description as a fallback — real GitHub bios land in a future update.