sunholo-data

Run ONE outer-loop iteration of a long-running mission (default: the V1 mission) — observe mission state, pick the top backlog item, route it through design-doc-creator → sprint-planner → sprint-executor → sprint-evaluator with the mission's model routing policy, record a log entry, and run the retro. Use when user says "run mission control", "mission iteration", "work the v1 backlog", or when fired nightly by the dev.ailang.mission-control launchd job.

Create AILANG design documents in the correct format and location. Use when user asks to create a design doc, plan a feature, or document a design. Handles both planned/ and implemented/ docs with proper structure.

Execute approved sprint plans with test-driven development, continuous linting, progress tracking, and pause points. Supports parallel milestone execution via Task sub-agents. Use when user says "execute sprint", "start sprint", or wants to implement an approved sprint plan.

Triage SonarCloud issues and hotspots via the SonarCloud Web API — check gate status, list open BLOCKER/CRITICAL bugs + vulns, group hotspots by rule, bulk-mark hotspots Safe per rule, transition issues to False Positive, and change the New Code period. Use whenever the user mentions SonarCloud, Sonar, quality gate red/failing, code smells to triage, marking hotspots safe, marking issues as false positive, setting new code period, or asks "why is the Sonar gate red" — even if they only reference a specific rule like go:S2245 without naming SonarCloud.

Run automated post-release workflow (eval baselines, dashboard, docs) for AILANG releases. Runs the tier-based benchmark suite (core+stretch+frontier by default) for standard + agent evals with validation and progress reporting. Use when user says "post-release tasks for vX.X.X" or "update dashboard". Fully autonomous with pre-flight checks.

Create and manage AILANG eval benchmarks. Use when user asks to create benchmarks, fix benchmark issues, debug failing benchmarks, or analyze benchmark results.

Create new AILANG releases with version bumps, changelog updates, git tags, and CI/CD verification. Use when user says "ready to release", "create release", mentions version numbers, or wants to publish a new version.

Evaluate sprint implementations against design docs and acceptance criteria. Runs after sprint-executor to assess quality with concrete scoring rubric (100 points, 70 to pass). Provides actionable feedback loop on failure. Use when user says "evaluate sprint", "review implementation", "assess sprint quality", or automatically via coordinator after sprint execution.

Write, review, debug, and explain code. Use when the user asks about programming, needs code written, wants a code review, or has a bug to investigate.

Extract structured data from documents, images, and unstructured text. Use when the user needs tables, key-value pairs, or specific fields pulled from content.

Analyze documents, extract structured data, and answer questions about uploaded files. Use when the user uploads a document, asks about file contents, or requests data extraction.

General-purpose AI assistant for everyday tasks. Use as the default skill when no specialized skill matches the user's request.

Search the web, summarize findings, and answer research questions. Use when the user asks about current events, needs web research, or wants information from online sources.

Interactive multi-surface A2UI demo. Renders a form to the workspace pane; user submits a value via a Button action; the agent reads the structured action context on the next turn without re-rendering. Demonstrates the discrete-action half of the v6.2.0 sprint 2.10 surface→agent loop — sibling of the read-only workspace-demo.

Multi-surface A2UI demo. Emits dashboard components to the persistent workspace pane instead of inline-in-chat. The deterministic end-to-end demo for the v6.2.0 sprint 2.9 surface routing.

MUST READ before writing or modifying any ADK agent code. ADK API quick reference for Python. Covers agent types, tool definitions, orchestration patterns, callbacks, state management, and common pitfalls. Use when user asks "how do I define an agent", "create a tool", "what orchestration patterns exist", "how do callbacks work", or when writing Agent(), FunctionTool, callbacks, or orchestration code. Do NOT use for deployment, evaluation, or project scaffolding questions.

Keep the AIPLA user-facing how-to guides (docs/guides/) up to date and in sync with the product. Use when the user says "update the guides", "regenerate / republish the guides", "the guides are stale / out of date", "re-capture the guide screenshots", "add a new guide", or mentions "guide-staleness" — and also proactively when a change lands on a UI surface a guide documents (the teacher class/activity/materials/co-pilot flows, the student join/workspace, or the researcher views). Covers the full pipeline: render, screenshot capture on deployed dev, publish into the app, seed the in-product corpus + onboarding tutors, and the staleness check.

In-app help assistant for teachers and researchers learning to use the AIPLA platform. Answers natural-language questions about setting up classes, building activities, adding curriculum materials, using the authoring co-pilot, the student experience, and the researcher views — grounded in the AIPLA how-to guides. Read-only: it explains how, it doesn't change anything. Visible to teachers and researchers (tagged role:teacher).

Teacher-facing co-pilot that helps a teacher author a good activity. The teacher describes, in plain Danish or English, what they want to teach; the assistant PROPOSES a Socratic lesson prompt (and, later, fitting elements) as editable suggestions the teacher accepts or edits — it never publishes on its own. Teacher-only (tagged role:teacher). Dark-flagged until the researcher teaching framework lands (designs 1.1.39 + 1.1.50).

Add a new MCP App artefact (hand-curated HTML/JS rendered as a sandboxed iframe in the chat workspace) to the AIPLA fork. Covers the static-artefact path (one shared mcp-sandbox Cloud Run service, files under infrastructure/mcp-sandbox/artefacts/<name>/v<version>/), the decision tree for "static vs. dynamic MCP server", the ADR-013 security gates (CSP, sandbox flags, 200 KB size limit, library-bypass review), and the frontend wiring (NEXT_PUBLIC_MCP_SANDBOX_URL, MCPAppToolCallRouter, workspace surface mount). Use when the user says "add a new MCP app", "new artefact", "build a sim", "deploy a sandbox iframe", "embed an interactive visualization", or references Boldkast / mcp-sandbox / mcp-ext-apps-map. Do NOT use for the agent-skills authoring workflow (that's the inherited template's skill creator) or for dynamic-MCP-server scaffolding (a separate v1 skill).

Teacher-facing analytics assistant. Ask natural-language questions about your class's session data — message counts, time-on-task, common misconceptions, group engagement. Only visible to teachers (tagged role:teacher). Data queries are scoped to the requesting teacher's own classes; no cross-teacher data exposure.

Teacher-facing hub for running your classes from chat — the conversational alternative to the React /teacher dashboard. Create a class, list your classes, mint group join-codes, look up your activities, and ask about a class's engagement (it consults the analytics assistant for session-data questions). Same backend, one chat. Only visible to teachers (tagged role:teacher).

Wire a new student-facing workbench/activity element so it reaches the tutor on BOTH surfaces — the iframe-context state push AND the visible "shared with the AI" human-tool-use trust card — AND make it proposable by the activity-authoring co-pilot (the add_element tool). Use when adding a workbench element, a new activity element, when a student interaction (table/calculator/checklist/etc.) does not show in the chat, when "the AI didn't see what the student did", when wiring or debugging the trust card, when a new element isn't offered by the authoring co-pilot, or reviewing an element PR for the dropped-card or dropped-co-pilot-coverage bug.

Triage runbook for AIPLA's dependency security pile. Use whenever the CI gate's `security-audit` job fails on a PR, when the Monday weekly rolling issue surfaces a new CVE, or when a developer says "run the security audit", "check dependabot", "vulnerability sweep", "triage the security pile", or "is this gate going to pass". Encodes the reachability rubric (direct prod / transitive / dev-only / deprecated) developed during the 2026-06-05 20-alert sweep, plus the per-ecosystem commands (npm audit, pip-audit via uvx, dependabot via gh api) and the npm-override conflict pattern. The policy of record is docs/design/aipla/v1.1.0-feedback/security-monitoring-pipeline.md — this skill is the human-facing operationalisation of it.

Manage starmap data assets for Stapledons Voyage. Use when downloading star catalogs, galactic backgrounds, or processing astronomical data. (project)

Generate and manage game image assets using AI. Use when user asks to create sprites, textures, backgrounds, portraits, or other visual assets for the game. Handles 3D textures, style consistency, and manifest updates. (project)

Execute approved sprint plans for Stapledons Voyage using AILANG. ALL game logic must be AILANG - engine is rendering only. Use when user says "execute sprint", "start implementation", or wants to begin an approved sprint plan.

Validate codebase architecture and organization before releases. Use when user asks to 'check architecture', 'validate structure', 'pre-release check', or before versioning.

Plan development sprints for Stapledons Voyage game features. Analyzes design docs, estimates effort considering AILANG constraints, and creates realistic sprint plans. Use when user asks to "plan sprint" or estimate game feature timelines.

Run CLI dev tools for game development (world inspection, benchmarks, assets, simulation). Use when user wants to 'check performance', 'inspect world', 'validate assets', 'stress test', or asks about new CLI features.

Run visual tests, compare golden files, and report bugs for Stapledons Voyage. Use when user asks to run tests, check golden files, or report visual regressions.

Create design documents for Stapledons Voyage game features. Use when user asks to create a design doc, plan a feature, or document game mechanics. Handles both planned/ and implemented/ docs.

End-to-end checklist for adding a new input or output format to AILANG Parse (docparse). Use when the user says 'add support for X format', 'wire up a new parser', 'add .foo format', 'add a parser for .bar', 'support .baz files', 'ship format X', 'can we parse .qux', 'new format rollout', or mentions a file extension that isn't yet supported (e.g. 'what about RTF', 'add Jupyter notebooks'). Covers parser code, ailang.toml exports, WASM demo, vendor-wasm-packages.sh, homepage, workbench, all format-list surfaces across the docs site, SDK READMEs, release tag, registry publish, and the ailang messages send docparse downstream notification. Use even when only part of this is requested — the skill tells you which surfaces you'd otherwise miss.

Create a new AILANG Parse documentation/landing page targeting a specific keyword or topic. Use when user says 'new landing page', 'new page for X', 'create a page about X', 'landing page for keyword X', or wants to add a documentation page to the docs/ site. Also use when the user references long-tail keywords, SEO pages, or competitor comparison pages.

Run OfficeDocBench evaluation and refresh benchmark scores across the AILANG Parse website. Use when the user says 'run benchmarks', 'rerun the benchmark', 'refresh benchmark numbers', 'update bench scores', 'regenerate summary.json', mentions OfficeDocBench, asks to evaluate parsers, asks why scores are out of sync on the docs, or after any change to docparse/services/*.ail or benchmarks/officedocbench/scoring.py — even if they don't explicitly say 'benchmark'. Also use after editing any AILANG parser to verify no regression.

Audit the design_docs/ folder structure against the current project version. Use when user says 'audit design docs', 'check design docs', 'are design docs organized', 'verify design doc placement', 'tidy design docs', or asks about misplaced/stale design documents. Also use when the user asks to reorganize planned vs implemented docs, or after a version bump.

Fix a parser gap identified in the spec coverage audit. Use when user says 'fix a gap', 'implement X parsing', 'improve parser coverage', 'close a gap', 'next P1', or references a specific ECMA-376 spec section like '§17.9'. Also use when the user asks to improve DOCX/XLSX/PPTX parsing quality, add support for a missing Office feature, or mentions numbering, styles, hyperlinks, formatting, fields, merged cells, equations, or speaker notes in the context of parser improvement.

Verify generated Office documents (DOCX, PPTX, XLSX, ODT, ODP, ODS, HTML) are valid and open correctly. Use when: user asks to verify generated files, check document output quality, validate Office files, run the verification loop, test generated documents, check if files open in Keynote/Word/Numbers, or after changing any generator code (docx_generator, pptx_generator, xlsx_generator, odt_generator, odp_generator, ods_generator, html_generator). Also use when user says "verify", "check the files", "do they open", "run verification", or "test the output".

Conversational front door for the Accounts-Payable workflow. Chats with the user about invoices and AP policy, then transfers to the deterministic `ap-pipeline` SequentialAgent when an invoice needs processing. Never reproduces extracted invoice fields itself — the pipeline owns that.

Invoice field-extraction specialist for Accounts Payable. Reads the already-parsed content of an invoice document and returns clean, typed business fields (vendor, line items, totals, GL codes). Operates on a second layer above the AILANG Parse / DocParse step that runs at ingest time — does NOT re-parse the document.

Action specialist for Accounts Payable. Takes a validated invoice and either posts it to the ERP ledger or routes it to a human approval queue, then emits the final Invoice Review Card to the workspace pane. Final stage of the `ap-pipeline` SequentialAgent.

Grounded validation specialist for Accounts Payable. Checks extracted invoice fields against the vendor master, open purchase orders, and the tax/approval policy indexed in Vertex AI Search. Detects duplicates, PO mismatches, out-of-policy line items, and incorrect tax. Use after extraction, before any posting decision.

Deterministic Accounts-Payable processing pipeline. Walks Extract → Validate → Post in sequence via ADK's SequentialAgent. No LLM at the workflow level — the order is code, not prose. Use as a sub-skill of ap-orchestrator; not intended for direct user invocation.

Write AILANG code. ALWAYS run 'ailang prompt' first - it contains the current syntax rules and templates.

Cross-agent communication system with semantic search and GitHub sync. Check messages, find similar content, deduplicate, and sync with GitHub Issues for AI workflows across sessions.

Debug AILANG code errors. Use when you encounter type errors, parse errors, or runtime failures in AILANG programs.

Build and refine websites from user briefs received via ailang messages. Use when the website-builder inbox has unread messages, or when the user asks to build/update a website.

Manage releases for Python projects. Use when user asks to release, bump version, tag, publish to PyPI, or monitor release workflows. Also use for /release.

Build a new workshop demo in this app from A2UI (declarative JSON UI) or MCP Apps (sandboxed iframe widgets). This is the right skill for ANY request to create a new demo, skill, or playground built from those protocols. Two targets: (1) a real chat skill that appears at /chat/@workshop-user/<slug> and emits A2UI or renders an MCP App widget live in the product chat (the default — needs a Gemini key), and (2) a no-key /dev fixture page for offline teaching. Use whenever the user wants to add, create, make, scaffold, or build a demo, chat skill, playground, A2UI surface, or MCP App widget — including "make me a demo/skill that does X", "add a skill that shows up in chat", "a form in chat from A2UI", "add an MCP App widget", or "new /dev playground". NOT for AG-UI, file-browser, or rich-media (already-built fixtures).

Audit the LinkedIn sketch rubric — measure per-signal fire rates across rendered sketches, identify dead/collapsed/healthy signals, propose changes following the "every signal maps to an AILANG primitive" rule, apply edits to sketch_rubric_signals.ail + sketch_rubric.ail + tests + design doc, then re-seed memory-safely. Use when user asks to improve rubric scores, audit signal fire rates, analyse why sketches score low/high, add new signals, broaden detection patterns, or tune the rubric for the LinkedIn sketches demo.

Build presentation slide decks as self-contained HTML files with dark/light themes, keyboard navigation, and animated transitions. Use when user asks to create slides, build a presentation, make a slide deck, add slides to a presentation, or wants conference-ready visual assets. Also use when user says "new slides", "presentation for", or references an existing presentation folder.

Parse documents with the AILANG Parse API. Use when user asks to parse, extract, or convert documents (DOCX, PDF, PPTX, XLSX, CSV, HTML, Markdown, EPUB, EML, images, audio, video). Also use when user mentions AILANG Parse, document parsing, unstructured data extraction, or needs to work with Office files programmatically. Triggers on "parse this file", "extract text from", "convert document", "AILANG Parse", or any document format processing task.