{"id":"65a5d178-2c3e-4148-bcc5-d7f9081c326a","shortId":"h5fT96","kind":"skill","title":"understanding-platform","tagline":"Explains Altertable platform concepts and architecture. Use when asking what Altertable is or how agents, discoveries, memories, insights, and dashboards fit together.","description":"# Understanding Platform\n\n## Quick Start\n\nUse this framing when someone asks what Altertable is:\n\n1. Altertable is an operational data platform built for continuous, agent-driven workloads.\n2. Its lakehouse foundation combines real-time ingestion, fast columnar analytics, and open standards.\n3. Agents run continuously on top of this data layer to monitor, model, and analyze data.\n4. The platform's core operating loop is: `Insights/Dashboards -> Agent Monitoring -> Discoveries -> Human Review -> Memories -> Better Future Analysis`.\n\n## When to Use This Skill\n\n- User asks \"what is Altertable?\" or \"how does the platform work?\"\n- User wants the difference between agents, discoveries, and memories\n- User asks how insights and dashboards connect to monitoring\n- User asks how Altertable differs from traditional warehouse-first stacks\n- User needs a conceptual architecture explanation before implementation details\n\n## Core Platform Narrative\n\nMost data stacks were optimized for batch pipelines, dashboards, and occasional human queries. Altertable is optimized for always-on analysis where agents continuously consume data.\n\nUse these points in order:\n\n1. **Foundation:** modern lakehouse architecture with warehouse-grade performance and better economics for high query volume.\n2. **Access:** data stays continuously queryable by both humans and software.\n3. **Intelligence layer:** agents orchestrate multiple LLMs through an asynchronous job system.\n4. **Operational output:** discoveries surface anomalies, trends, and opportunities for human review.\n5. **Learning loop:** memories retain validated context and improve future agent behavior.\n\n## Concept Map\n\n### Agents\n\nAutonomous data collaborators that execute both repetitive and higher-level analytics work.\n\n- Synchronize sources and maintain data readiness\n- Build or update models, queries, and visual outputs\n- Monitor insights and dashboards continuously\n- Generate discoveries when something noteworthy happens\n- Learn from feedback through memories\n\n### Discoveries\n\nReviewable findings generated by agents.\n\n- Include context, rationale, and suggested actions\n- Require human approval or rejection\n- Can represent anomalies, trend changes, segment shifts, schema/model changes, and event readiness\n- Become a primary collaboration interface between agents and teams\n\n### Memories\n\nPersistent knowledge accumulated by agents across runs.\n\n- Episodic: what happened\n- Semantic: what it means\n- Procedural: how to handle it next time\n- Reinforced or weakened by discovery review outcomes and repeated use\n\n### Insights\n\nPersistent analyses and visualizations over lakehouse data.\n\n- Funnel, segmentation, semantic, and SQL insights cover different analysis needs\n- Serve as reusable analytical building blocks\n- Can be monitored directly by agents\n\n### Dashboards\n\nCollections of insights organized for KPI tracking and shared monitoring.\n\n- Aggregate related metrics and context in one place\n- Support shared variables for coordinated filtering\n- Can have attached agents that watch for anomalies and trend shifts\n\n## How the Concepts Work Together\n\n```text\nData ingestion -> Lakehouse storage/query engine\n    -> Insights and Dashboards\n        -> Continuous Agent Monitoring\n            -> Discoveries\n                -> Human Review (accept/reject)\n                    -> Memories updated\n                        -> Better future monitoring and analysis\n```\n\n## Communication Guidelines\n\nWhen explaining the platform:\n\n- Start with outcomes (continuous analysis, faster decisions, lower marginal cost at scale)\n- Then map to concepts (agents, discoveries, memories, insights, dashboards)\n- Emphasize human-in-the-loop review for quality and trust\n- Distinguish \"analysis artifacts\" (insights/dashboards) from \"agent outputs\" (discoveries)\n- Describe memories as adaptive context, not static storage\n\n## Common Pitfalls\n\n- Presenting Altertable as only a BI/dashboard tool\n- Describing agents as one-shot assistants instead of continuous collaborators\n- Skipping the human review stage in the discovery lifecycle\n- Treating discoveries as equivalent to insights (they are not)\n- Omitting memory feedback loops when explaining how agent quality improves over time\n- Leading with implementation internals before clarifying conceptual flow\n\n## Reference Files\n\n- [Platform overview](references/platform-overview.md)\n- [Agents](references/agents.md)\n- [Discoveries](references/discoveries.md)\n- [Memories](references/memories.md)\n- [Insights and dashboards](references/insights-and-dashboards.md)","tags":["understanding","platform","skills","altertable-ai","agent-skills","ai-agents","altertable"],"capabilities":["skill","source-altertable-ai","skill-understanding-platform","topic-agent-skills","topic-ai-agents","topic-altertable"],"categories":["skills"],"synonyms":[],"warnings":[],"endpointUrl":"https://skills.sh/altertable-ai/skills/understanding-platform","protocol":"skill","transport":"skills-sh","auth":{"type":"none","details":{"cli":"npx skills add altertable-ai/skills","source_repo":"https://github.com/altertable-ai/skills","install_from":"skills.sh"}},"qualityScore":"0.453","qualityRationale":"deterministic score 0.45 from registry signals: · indexed on github topic:agent-skills · 7 github stars · SKILL.md body (4,566 chars)","verified":false,"liveness":"unknown","lastLivenessCheck":null,"agentReviews":{"count":0,"score_avg":null,"cost_usd_avg":null,"success_rate":null,"latency_p50_ms":null,"narrative_summary":null,"summary_updated_at":null},"enrichmentModel":"deterministic:skill-github:v1","enrichmentVersion":1,"enrichedAt":"2026-05-18T19:14:20.871Z","embedding":null,"createdAt":"2026-05-18T13:21:55.631Z","updatedAt":"2026-05-18T19:14:20.871Z","lastSeenAt":"2026-05-18T19:14:20.871Z","tsv":"'1':39,190 '2':53,207 '3':68,218 '4':84,230 '5':242 'accept/reject':456 'access':208 'accumul':341 'across':344 'action':311 'adapt':513 'agent':18,50,69,93,123,181,221,252,256,305,335,343,399,428,451,486,507,528,563,581 'agent-driven':49 'aggreg':411 'altert':5,14,37,40,111,139,172,521 'alway':177 'always-on':176 'analys':372 'analysi':101,179,386,463,474,503 'analyt':64,268,391 'analyz':82 'anomali':235,319,432 'approv':314 'architectur':9,151,194 'artifact':504 'ask':12,35,108,128,137 'assist':533 'asynchron':227 'attach':427 'autonom':257 'batch':165 'becom':329 'behavior':253 'better':99,201,459 'bi/dashboard':525 'block':393 'build':276,392 'built':46 'chang':321,325 'clarifi':573 'collabor':259,332,537 'collect':401 'columnar':63 'combin':57 'common':518 'communic':464 'concept':7,254,438,485 'conceptu':150,574 'connect':133 'consum':183 'context':248,307,415,514 'continu':48,71,182,211,288,450,473,536 'coordin':423 'core':88,156 'cost':479 'cover':384 'dashboard':23,132,167,287,400,449,490,589 'data':44,76,83,160,184,209,258,274,377,442 'decis':476 'describ':510,527 'detail':155 'differ':121,140,385 'direct':397 'discoveri':19,95,124,233,290,300,364,453,487,509,545,548,583 'distinguish':502 'driven':51 'econom':202 'emphas':491 'engin':446 'episod':346 'equival':550 'event':327 'execut':261 'explain':4,467,561 'explan':152 'fast':62 'faster':475 'feedback':297,558 'file':577 'filter':424 'find':302 'first':145 'fit':24 'flow':575 'foundat':56,191 'frame':32 'funnel':378 'futur':100,251,460 'generat':289,303 'grade':198 'guidelin':465 'handl':356 'happen':294,348 'high':204 'higher':266 'higher-level':265 'human':96,170,215,240,313,454,493,540 'human-in-the-loop':492 'implement':154,570 'improv':250,565 'includ':306 'ingest':61,443 'insight':21,130,285,370,383,403,447,489,552,587 'insights/dashboards':92,505 'instead':534 'intellig':219 'interfac':333 'intern':571 'job':228 'knowledg':340 'kpi':406 'lakehous':55,193,376,444 'layer':77,220 'lead':568 'learn':243,295 'level':267 'lifecycl':546 'llms':224 'loop':90,244,496,559 'lower':477 'maintain':273 'map':255,483 'margin':478 'mean':352 'memori':20,98,126,245,299,338,457,488,511,557,585 'metric':413 'model':80,279 'modern':192 'monitor':79,94,135,284,396,410,452,461 'multipl':223 'narrat':158 'need':148,387 'next':358 'noteworthi':293 'occasion':169 'omit':556 'one':417,531 'one-shot':530 'open':66 'oper':43,89,231 'opportun':238 'optim':163,174 'orchestr':222 'order':189 'organ':404 'outcom':366,472 'output':232,283,508 'overview':579 'perform':199 'persist':339,371 'pipelin':166 'pitfal':519 'place':418 'platform':3,6,27,45,86,116,157,469,578 'point':187 'present':520 'primari':331 'procedur':353 'qualiti':499,564 'queri':171,205,280 'queryabl':212 'quick':28 'rational':308 'readi':275,328 'real':59 'real-tim':58 'refer':576 'references/agents.md':582 'references/discoveries.md':584 'references/insights-and-dashboards.md':590 'references/memories.md':586 'references/platform-overview.md':580 'reinforc':360 'reject':316 'relat':412 'repeat':368 'repetit':263 'repres':318 'requir':312 'retain':246 'reusabl':390 'review':97,241,301,365,455,497,541 'run':70,345 'scale':481 'schema/model':324 'segment':322,379 'semant':349,380 'serv':388 'share':409,420 'shift':323,435 'shot':532 'skill':106 'skill-understanding-platform' 'skip':538 'softwar':217 'someon':34 'someth':292 'sourc':271 'source-altertable-ai' 'sql':382 'stack':146,161 'stage':542 'standard':67 'start':29,470 'static':516 'stay':210 'storag':517 'storage/query':445 'suggest':310 'support':419 'surfac':234 'synchron':270 'system':229 'team':337 'text':441 'time':60,359,567 'togeth':25,440 'tool':526 'top':73 'topic-agent-skills' 'topic-ai-agents' 'topic-altertable' 'track':407 'tradit':142 'treat':547 'trend':236,320,434 'trust':501 'understand':2,26 'understanding-platform':1 'updat':278,458 'use':10,30,104,185,369 'user':107,118,127,136,147 'valid':247 'variabl':421 'visual':282,374 'volum':206 'want':119 'warehous':144,197 'warehouse-first':143 'warehouse-grad':196 'watch':430 'weaken':362 'work':117,269,439 'workload':52","prices":[{"id":"4ceb6fb2-bfac-4e53-8e5f-f3c045c055d8","listingId":"65a5d178-2c3e-4148-bcc5-d7f9081c326a","amountUsd":"0","unit":"free","nativeCurrency":null,"nativeAmount":null,"chain":null,"payTo":null,"paymentMethod":"skill-free","isPrimary":true,"details":{"org":"altertable-ai","category":"skills","install_from":"skills.sh"},"createdAt":"2026-05-18T13:21:55.631Z"}],"sources":[{"listingId":"65a5d178-2c3e-4148-bcc5-d7f9081c326a","source":"github","sourceId":"altertable-ai/skills/understanding-platform","sourceUrl":"https://github.com/altertable-ai/skills/tree/main/skills/understanding-platform","isPrimary":false,"firstSeenAt":"2026-05-18T13:21:55.631Z","lastSeenAt":"2026-05-18T19:14:20.871Z"}],"details":{"listingId":"65a5d178-2c3e-4148-bcc5-d7f9081c326a","quickStartSnippet":null,"exampleRequest":null,"exampleResponse":null,"schema":null,"openapiUrl":null,"agentsTxtUrl":null,"citations":[],"useCases":[],"bestFor":[],"notFor":[],"kindDetails":{"org":"altertable-ai","slug":"understanding-platform","github":{"repo":"altertable-ai/skills","stars":7,"topics":["agent-skills","ai-agents","altertable"],"license":"mit","html_url":"https://github.com/altertable-ai/skills","pushed_at":"2026-05-14T10:34:10Z","description":"Agent Skills for Altertable","skill_md_sha":"105135cddc0c741b9c766576e711abe054078d01","skill_md_path":"skills/understanding-platform/SKILL.md","default_branch":"main","skill_tree_url":"https://github.com/altertable-ai/skills/tree/main/skills/understanding-platform"},"layout":"multi","source":"github","category":"skills","frontmatter":{"name":"understanding-platform","description":"Explains Altertable platform concepts and architecture. Use when asking what Altertable is or how agents, discoveries, memories, insights, and dashboards fit together.","compatibility":"Requires Altertable MCP server"},"skills_sh_url":"https://skills.sh/altertable-ai/skills/understanding-platform"},"updatedAt":"2026-05-18T19:14:20.871Z"}}