[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M77JHBQF0S4W66Y8GJFSJS42":3,"skill-same:01M77JHBQF0S4W66Y8GJFSJS42":45},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":15,"categories":16,"keywords":17,"publishedAt":18,"linkOnly":19,"markdownContent":20,"markdownBody":21,"frontMatter":22,"refName":23,"commitSha":24,"skillDirectory":25,"forksCount":26,"repositoryHtmlUrl":27,"repositoryHomepageUrl":28,"ownerHtmlUrl":29,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":30,"install":31,"resources":44},"01M77JHBQF0S4W66Y8GJFSJS42","powerbi-workflows",null,"Route guarded Power BI work -- design, native report authoring, semantic-model operations, published queries, QA, bounded formatting, and PBIP adoption -- to the correct Seshat or official Microsoft surface under Seshat BI's gates.","Kemetra\u002FSeshat-BI","Kemetra","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F206601658?v=4","01M77J5APND9QAJK8XMBH5B1A8","01M77J5AR4TM1GN2G4GS5KH02J",2,"Apache-2.0","https:\u002F\u002Fgithub.com\u002FKemetra\u002FSeshat-BI\u002Fblob\u002Fmain\u002Fdistribution\u002Fbundle-templates\u002Fshared\u002Fskills\u002Fpowerbi-workflows\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: powerbi-workflows\ndescription: >-\n  Route guarded Power BI work -- design, native report authoring, semantic-model\n  operations, published queries, QA, bounded formatting, and PBIP adoption --\n  to the correct Seshat or official Microsoft surface under Seshat BI's gates.\n---\n\n# Power BI workflows\n\nRead `..\u002F..\u002Fportable-operating-contract.md` before acting. These routes never\nreplace readiness gates: no invented metric, measure, KPI, or DAX meaning; no\nnumeric readiness\u002Fconfidence score; no self-granted approval; no\ndashboard-ready claim without committed evidence.\n\n## One front door, explicit execution owners\n\nThis skill is the broad Power BI front door. It owns intent classification and\nSeshat's pre\u002Fpost gates; it does not reproduce Microsoft execution behavior.\nFor execution-shaped requests, use `seshat pbi-mcp doctor` as the canonical\nmachine-checkable selector even when the selected owner is an official skill\nrather than an MCP server.\n\n| User intent | Seshat pre-gate | Execution owner | Seshat after execution |\n|---|---|---|---|\n| Business\u002Freport intent and metric meaning | approved decisions and contracts | Seshat knowledge\u002Fgovernance | readiness and evidence |\n| Dashboard\u002Fpage design | approved metrics, semantic evidence, narrative brief | Microsoft `powerbi-report-design` | Seshat human-review and evidence gates |\n| Native PBIR page\u002Fvisual\u002Ffilter\u002Fslicer\u002Fbinding authoring | exact target semantic pass, named-human `dashboard_ready` approval, official skill discoverable | Microsoft `powerbi-report-authoring` | binding, blueprint, and static validation |\n| Bounded theme\u002Fformat\u002Fbackground\u002Fgeometry edits | exact target semantic pass, named-human `dashboard_ready` approval, and command allow-list | temporary Seshat PBIR gap adapter | binding-preservation and static validation |\n| Semantic-model edit | exact target semantic pass; F016 policy | BLOCKED while F016 is parked | no execution |\n| Published semantic-model query | governed target and tenant prerequisites | Microsoft remote Power BI MCP | evidence interpretation |\n| PBIP inspection\u002Fadoption | repository target | Seshat read-only tools | findings and next action |\n\nFor native report authoring run:\n\n`seshat pbi-mcp doctor --repo . --intent report-authoring --target \u003Ctable> --harness \u003Cclaude-code|codex>`\n\nSpec 148 provides the read-only harness proof. Before delegation, run\n`seshat integrations setup --profile powerbi-fabric --harness \u003Cclaude-code|codex>`\nand require the `fabric-skills` harness result to be discoverable. Never treat a\ncloned bundle as an activated executor, and never copy its implementation into\nSeshat. The current full Claude Power BI plugin is incompatible because it also\nactivates planning, management, semantic-authoring, and an unsafe modeling-MCP\nsurface; do not selectively ignore those extras. Compatible Codex projections\nremain exact-skill, provenance-checked activations.\n\nThe named-human approval authorizes authoring but does not self-grant or require\n`dashboard_ready: pass`; that status can follow only after the resulting report\npasses the governed review and validation sequence.\n\n## Design (dashboards and pages)\n\nCheck the proposal first with the installed read-only helpers when available:\n`seshat dashboard-planner` returns a categorical new\u002Fextends\u002Fduplicate verdict\nagainst the committed dashboard set (`--proposal`\u002F`--tuple`), and\n`seshat dashboard-gaps` inventories design-blocking gaps before any layout\nwork. The `dashboard-gaps` `--page-intent` file is a YAML mapping with a\n`questions:` list (each question naming its required metrics and dimensions);\nstart from the page-intent example the kit ships and see the\ndashboard-gap-detector guide for the shape. A missing, unreadable,\nor wrong-format page-intent is refused with a named error and exit 2 -- it is\nnever an empty \"no gaps\" read.\n\nData-bound visual design requires approved metric contracts, committed\nsemantic-model evidence, AND a committed narrative brief. The narrative brief\n(`mappings\u002F\u003Ctable>\u002Fnarrative-brief.md`, frozen `seshat.narrative-brief\u002Fv1`\nschema) is the NEW narrative gate: a design that binds visuals to contracts but\ncannot say which owner decision each page serves is itself a gap. Author the\nbrief FIRST via the `bi-analyst-knowledge` derivation route (ranked\ndecision-questions, one framing per question, a story order, honest `[GAP]`\nentries); its absence is a named blocker, not a warning -- stop and name it.\n\nWith all three gates present, produce reviewable design guidance -- a layout\nplan, a visual list, and a THREE-WAY binding map (visual -> contract ->\ndecision-question) where every data-bound visual binds to exactly one approved\ncontract AND answers at least one brief decision-question. An orphan in either\ndirection is a defect; every declared page must serve at least one decision; and\nevery headline (KPI-card class) visual must answer a `stage: overview` question\nthat names a comparison -- a bare total on a headline is a defect. Give the map a\nmachine-readable `seshat.binding-map\u002Fv1` front section and check it read-only\nwith `seshat narrative-check --table \u003Ct> --binding-map` (and the brief with\n`seshat narrative-check --table \u003Ct>`); a clean check is evidence for the human\nreview, never an approval. Slicers and filters belong to a compact filter rail\nthat never dominates the canvas, and each slicer's field and default selection is\npart of the reviewable design. Without the gates, stop and name the missing one.\nFor the analyst framing catalog + derivation route load `bi-analyst-knowledge`;\nfor metric meaning load `retail-kpi-knowledge`; for measure semantics load\n`bi-dax-knowledge`.\n\n## Review and QA\n\nReview a screenshot or built report against the design guidance above and\nreport concrete, advisory findings. Validate a page blueprint with the\ninstalled `seshat pbir-validate-blueprint` helper when available. Before a\nhuman opens Desktop on any agent-touched report, run the installed\n`seshat pbir-validate-bindings --report \u003CX.Report> --model \u003CX.SemanticModel>`\nhelper when available: it resolves every bound field (projections, filters,\nsorts) against the model's TMDL and blocks on unresolved bindings -- missing\nmeasures\u002Fcolumns, unknown entities, PII-masked renames -- the exact class that\notherwise surfaces as Desktop error cards. It needs no blueprint or binding\nmap, so it also covers Desktop-owned reports. A clean review is evidence for a\nnamed human, never an approval.\n\nA clean binding report does NOT mean the model loads. It checks BINDINGS only;\nit does not verify TMDL syntax, so a model with a TMDL defect can pass it and\nstill fail to open in Desktop. Never report \"validated\" or \"Desktop-ready\" on\nthe strength of this check alone -- say bindings resolve, and name TMDL\nloadability as unverified.\n\nWhen you have hand-authored TMDL, the installed\n`seshat tmdl-doc-comment-lint --model \u003CX.SemanticModel>` helper catches ONE\nmistake that is otherwise invisible until Desktop: a `\u002F\u002F\u002F` documentation block\nfollowed by a blank line instead of the declaration it documents, which makes\nDesktop reject the whole project. It checks that one rule and nothing else --\nit is NOT a TMDL syntax validator, so a pass does NOT mean the TMDL is valid or\nthat Desktop can load the model. TMDL loadability stays unverified either way;\nrunning both helpers narrows two known classes, it does not clear the model.\n\n## Reopening Desktop after an external edit\n\nPower BI Desktop does not re-read PBIR\u002FTMDL files edited on disk: a running\nDesktop serves its in-memory session, and a reopen via File > Open or the\nRecent list restores cached view state from `.pbi\u002FlocalSettings.json` -- the\non-disk edits stay invisible and the old layout (plus phantom error cards) can\npersist across close\u002Freopen cycles. After any agent-authored write to a\nDesktop-owned project, follow this protocol before a human looks at the\nreport:\n\n1. Fully quit Desktop and verify both `PBIDesktop.exe` and `msmdsrv.exe` are\n   gone -- closing the report tab is not enough.\n2. Bump the modification times of the edited `definition\u002F` files so they are\n   newer than Desktop's last-known state.\n3. Move `.pbi\u002FlocalSettings.json` aside for BOTH the `.Report` and the\n   `.SemanticModel` folders (Desktop regenerates it, with no stale view state\n   left to restore).\n4. Reopen the project by double-clicking the `.pbip` in File Explorer, not\n   from Desktop's Recent list.\n5. If Desktop prompts to save a stale session, choose \"Don't Save\" so the old\n   in-memory layout does not overwrite the good on-disk edits.\n\n## Theme and backgrounds\n\nGenerate theme artifacts with `seshat theme-gen` and `seshat theme-compile`;\napply them to a committed report with\n`seshat pbir-apply-theme --repo . --table \u003Ctable>` and\n`seshat pbir-set-page-background --repo . --table \u003Ctable>`. Themes cover palette, fonts, visual\ndefaults, page\u002Fwallpaper defaults, sentiment colors, and filter-pane\u002F\nfilter-card defaults -- the filter pane's LOOK, never what it filters. Theme\nand background files carry style and structure only -- never business data,\nmetric meaning, secrets, or PII.\n\n## Formatting and geometry\n\nAuthor formatting plans freely, but mutate committed PBIR only through the\nallow-listed installed helpers\n`seshat pbir-format-visual --repo . --table \u003Ctable>` and\n`seshat pbir-set-geometry --repo . --table \u003Ctable>`, which preserve every data\nbinding byte-for-byte. These four helpers are the temporary reviewed\n`powerbi-bounded-pbir-patching` gap, not general report authoring.\nAdding a slicer or changing what a visual or filter binds to is a BINDING\nchange, not formatting -- route it back to the design gate. Anything outside\nthe allow-list routes to official `powerbi-report-authoring` after the design\ngate; if that skill is not discoverable, stop and report the integration gap.\n\n## Semantic measures (handoff)\n\nFrom an APPROVED metric contract only, the installed `seshat generate\n--contract \u003Cpath>` produces a verified TMDL measure block into a new\nstandalone file (never under a `powerbi\u002F` tree, never overwriting); it does\nnot invent meaning beyond the contract. With a live database and an\nowner-approved expected value, `seshat value-check` compares a measure's\nrecomputed aggregate within tolerance -- report a pending state if the\ndatabase extra or DSN is absent.\n\n## Existing PBIP adoption\n\nFollow the `seshat-bi` skill's adoption route: read-only\n`seshat adopt-pbip assess`, human review of the exact assessment digest, then\n`seshat adopt-pbip scaffold` in a clean Git worktree.\n\nIf `seshat` is unavailable, explain that the Python package `seshat-bi` must be\ninstalled; report a pending state rather than simulating helper output.\n","# Power BI workflows\n\nRead `..\u002F..\u002Fportable-operating-contract.md` before acting. These routes never\nreplace readiness gates: no invented metric, measure, KPI, or DAX meaning; no\nnumeric readiness\u002Fconfidence score; no self-granted approval; no\ndashboard-ready claim without committed evidence.\n\n## One front door, explicit execution owners\n\nThis skill is the broad Power BI front door. It owns intent classification and\nSeshat's pre\u002Fpost gates; it does not reproduce Microsoft execution behavior.\nFor execution-shaped requests, use `seshat pbi-mcp doctor` as the canonical\nmachine-checkable selector even when the selected owner is an official skill\nrather than an MCP server.\n\n| User intent | Seshat pre-gate | Execution owner | Seshat after execution |\n|---|---|---|---|\n| Business\u002Freport intent and metric meaning | approved decisions and contracts | Seshat knowledge\u002Fgovernance | readiness and evidence |\n| Dashboard\u002Fpage design | approved metrics, semantic evidence, narrative brief | Microsoft `powerbi-report-design` | Seshat human-review and evidence gates |\n| Native PBIR page\u002Fvisual\u002Ffilter\u002Fslicer\u002Fbinding authoring | exact target semantic pass, named-human `dashboard_ready` approval, official skill discoverable | Microsoft `powerbi-report-authoring` | binding, blueprint, and static validation |\n| Bounded theme\u002Fformat\u002Fbackground\u002Fgeometry edits | exact target semantic pass, named-human `dashboard_ready` approval, and command allow-list | temporary Seshat PBIR gap adapter | binding-preservation and static validation |\n| Semantic-model edit | exact target semantic pass; F016 policy | BLOCKED while F016 is parked | no execution |\n| Published semantic-model query | governed target and tenant prerequisites | Microsoft remote Power BI MCP | evidence interpretation |\n| PBIP inspection\u002Fadoption | repository target | Seshat read-only tools | findings and next action |\n\nFor native report authoring run:\n\n`seshat pbi-mcp doctor --repo . --intent report-authoring --target \u003Ctable> --harness \u003Cclaude-code|codex>`\n\nSpec 148 provides the read-only harness proof. Before delegation, run\n`seshat integrations setup --profile powerbi-fabric --harness \u003Cclaude-code|codex>`\nand require the `fabric-skills` harness result to be discoverable. Never treat a\ncloned bundle as an activated executor, and never copy its implementation into\nSeshat. The current full Claude Power BI plugin is incompatible because it also\nactivates planning, management, semantic-authoring, and an unsafe modeling-MCP\nsurface; do not selectively ignore those extras. Compatible Codex projections\nremain exact-skill, provenance-checked activations.\n\nThe named-human approval authorizes authoring but does not self-grant or require\n`dashboard_ready: pass`; that status can follow only after the resulting report\npasses the governed review and validation sequence.\n\n## Design (dashboards and pages)\n\nCheck the proposal first with the installed read-only helpers when available:\n`seshat dashboard-planner` returns a categorical new\u002Fextends\u002Fduplicate verdict\nagainst the committed dashboard set (`--proposal`\u002F`--tuple`), and\n`seshat dashboard-gaps` inventories design-blocking gaps before any layout\nwork. The `dashboard-gaps` `--page-intent` file is a YAML mapping with a\n`questions:` list (each question naming its required metrics and dimensions);\nstart from the page-intent example the kit ships and see the\ndashboard-gap-detector guide for the shape. A missing, unreadable,\nor wrong-format page-intent is refused with a named error and exit 2 -- it is\nnever an empty \"no gaps\" read.\n\nData-bound visual design requires approved metric contracts, committed\nsemantic-model evidence, AND a committed narrative brief. The narrative brief\n(`mappings\u002F\u003Ctable>\u002Fnarrative-brief.md`, frozen `seshat.narrative-brief\u002Fv1`\nschema) is the NEW narrative gate: a design that binds visuals to contracts but\ncannot say which owner decision each page serves is itself a gap. Author the\nbrief FIRST via the `bi-analyst-knowledge` derivation route (ranked\ndecision-questions, one framing per question, a story order, honest `[GAP]`\nentries); its absence is a named blocker, not a warning -- stop and name it.\n\nWith all three gates present, produce reviewable design guidance -- a layout\nplan, a visual list, and a THREE-WAY binding map (visual -> contract ->\ndecision-question) where every data-bound visual binds to exactly one approved\ncontract AND answers at least one brief decision-question. An orphan in either\ndirection is a defect; every declared page must serve at least one decision; and\nevery headline (KPI-card class) visual must answer a `stage: overview` question\nthat names a comparison -- a bare total on a headline is a defect. Give the map a\nmachine-readable `seshat.binding-map\u002Fv1` front section and check it read-only\nwith `seshat narrative-check --table \u003Ct> --binding-map` (and the brief with\n`seshat narrative-check --table \u003Ct>`); a clean check is evidence for the human\nreview, never an approval. Slicers and filters belong to a compact filter rail\nthat never dominates the canvas, and each slicer's field and default selection is\npart of the reviewable design. Without the gates, stop and name the missing one.\nFor the analyst framing catalog + derivation route load `bi-analyst-knowledge`;\nfor metric meaning load `retail-kpi-knowledge`; for measure semantics load\n`bi-dax-knowledge`.\n\n## Review and QA\n\nReview a screenshot or built report against the design guidance above and\nreport concrete, advisory findings. Validate a page blueprint with the\ninstalled `seshat pbir-validate-blueprint` helper when available. Before a\nhuman opens Desktop on any agent-touched report, run the installed\n`seshat pbir-validate-bindings --report \u003CX.Report> --model \u003CX.SemanticModel>`\nhelper when available: it resolves every bound field (projections, filters,\nsorts) against the model's TMDL and blocks on unresolved bindings -- missing\nmeasures\u002Fcolumns, unknown entities, PII-masked renames -- the exact class that\notherwise surfaces as Desktop error cards. It needs no blueprint or binding\nmap, so it also covers Desktop-owned reports. A clean review is evidence for a\nnamed human, never an approval.\n\nA clean binding report does NOT mean the model loads. It checks BINDINGS only;\nit does not verify TMDL syntax, so a model with a TMDL defect can pass it and\nstill fail to open in Desktop. Never report \"validated\" or \"Desktop-ready\" on\nthe strength of this check alone -- say bindings resolve, and name TMDL\nloadability as unverified.\n\nWhen you have hand-authored TMDL, the installed\n`seshat tmdl-doc-comment-lint --model \u003CX.SemanticModel>` helper catches ONE\nmistake that is otherwise invisible until Desktop: a `\u002F\u002F\u002F` documentation block\nfollowed by a blank line instead of the declaration it documents, which makes\nDesktop reject the whole project. It checks that one rule and nothing else --\nit is NOT a TMDL syntax validator, so a pass does NOT mean the TMDL is valid or\nthat Desktop can load the model. TMDL loadability stays unverified either way;\nrunning both helpers narrows two known classes, it does not clear the model.\n\n## Reopening Desktop after an external edit\n\nPower BI Desktop does not re-read PBIR\u002FTMDL files edited on disk: a running\nDesktop serves its in-memory session, and a reopen via File > Open or the\nRecent list restores cached view state from `.pbi\u002FlocalSettings.json` -- the\non-disk edits stay invisible and the old layout (plus phantom error cards) can\npersist across close\u002Freopen cycles. After any agent-authored write to a\nDesktop-owned project, follow this protocol before a human looks at the\nreport:\n\n1. Fully quit Desktop and verify both `PBIDesktop.exe` and `msmdsrv.exe` are\n   gone -- closing the report tab is not enough.\n2. Bump the modification times of the edited `definition\u002F` files so they are\n   newer than Desktop's last-known state.\n3. Move `.pbi\u002FlocalSettings.json` aside for BOTH the `.Report` and the\n   `.SemanticModel` folders (Desktop regenerates it, with no stale view state\n   left to restore).\n4. Reopen the project by double-clicking the `.pbip` in File Explorer, not\n   from Desktop's Recent list.\n5. If Desktop prompts to save a stale session, choose \"Don't Save\" so the old\n   in-memory layout does not overwrite the good on-disk edits.\n\n## Theme and backgrounds\n\nGenerate theme artifacts with `seshat theme-gen` and `seshat theme-compile`;\napply them to a committed report with\n`seshat pbir-apply-theme --repo . --table \u003Ctable>` and\n`seshat pbir-set-page-background --repo . --table \u003Ctable>`. Themes cover palette, fonts, visual\ndefaults, page\u002Fwallpaper defaults, sentiment colors, and filter-pane\u002F\nfilter-card defaults -- the filter pane's LOOK, never what it filters. Theme\nand background files carry style and structure only -- never business data,\nmetric meaning, secrets, or PII.\n\n## Formatting and geometry\n\nAuthor formatting plans freely, but mutate committed PBIR only through the\nallow-listed installed helpers\n`seshat pbir-format-visual --repo . --table \u003Ctable>` and\n`seshat pbir-set-geometry --repo . --table \u003Ctable>`, which preserve every data\nbinding byte-for-byte. These four helpers are the temporary reviewed\n`powerbi-bounded-pbir-patching` gap, not general report authoring.\nAdding a slicer or changing what a visual or filter binds to is a BINDING\nchange, not formatting -- route it back to the design gate. Anything outside\nthe allow-list routes to official `powerbi-report-authoring` after the design\ngate; if that skill is not discoverable, stop and report the integration gap.\n\n## Semantic measures (handoff)\n\nFrom an APPROVED metric contract only, the installed `seshat generate\n--contract \u003Cpath>` produces a verified TMDL measure block into a new\nstandalone file (never under a `powerbi\u002F` tree, never overwriting); it does\nnot invent meaning beyond the contract. With a live database and an\nowner-approved expected value, `seshat value-check` compares a measure's\nrecomputed aggregate within tolerance -- report a pending state if the\ndatabase extra or DSN is absent.\n\n## Existing PBIP adoption\n\nFollow the `seshat-bi` skill's adoption route: read-only\n`seshat adopt-pbip assess`, human review of the exact assessment digest, then\n`seshat adopt-pbip scaffold` in a clean Git worktree.\n\nIf `seshat` is unavailable, explain that the Python package `seshat-bi` must be\ninstalled; report a pending state rather than simulating helper output.\n",{"name":5,"description":7},"main","6266e304556a72b75d951b9433fd833481c61233","distribution\u002Fbundle-templates\u002Fshared\u002Fskills\u002Fpowerbi-workflows",0,"https:\u002F\u002Fgithub.com\u002FKemetra\u002FSeshat-BI","https:\u002F\u002Fpypi.org\u002Fproject\u002Fseshat-bi\u002F","https:\u002F\u002Fgithub.com\u002FKemetra","2026-11-11 21:44:25",{"source":8,"skillName":5,"commands":32,"prompt":42,"zipAvailable":43},[33,36,39],{"tool":34,"command":35},"npx","npx skills add Kemetra\u002FSeshat-BI --skill powerbi-workflows",{"tool":37,"command":38},"bunx","bunx skills add Kemetra\u002FSeshat-BI --skill powerbi-workflows",{"tool":40,"command":41},"pnpm","pnpm dlx skills add Kemetra\u002FSeshat-BI --skill powerbi-workflows","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FKemetra\u002FSeshat-BI\u002Fblob\u002Fmain\u002Fdistribution\u002Fbundle-templates\u002Fshared\u002Fskills\u002Fpowerbi-workflows\u002FSKILL.md",true,[],[46,53,58,64,69,76],{"skillNo":47,"name":48,"nameZh":6,"description":49,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":50,"categories":51,"keywords":52,"publishedAt":18},"01M77JWCTJ88JWPDD25WZ2NEY8","fake-skill","A fixture skill with declaring frontmatter, used as positive shipped evidence.","https:\u002F\u002Fgithub.com\u002FKemetra\u002FSeshat-BI\u002Fblob\u002Fmain\u002Ftests\u002Funit\u002Ffixtures\u002Fcapability_inventory\u002Funlisted\u002F.claude\u002Fskills\u002Ffake-skill\u002FSKILL.md",[],[],{"skillNo":54,"name":48,"nameZh":6,"description":49,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":55,"categories":56,"keywords":57,"publishedAt":18},"01M77JWA0BTG4JB2HR52Q99216","https:\u002F\u002Fgithub.com\u002FKemetra\u002FSeshat-BI\u002Fblob\u002Fmain\u002Ftests\u002Funit\u002Ffixtures\u002Fcapability_inventory\u002Forphan\u002F.claude\u002Fskills\u002Ffake-skill\u002FSKILL.md",[],[],{"skillNo":59,"name":60,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":61,"categories":62,"keywords":63,"publishedAt":18},"01M77JW6VE59XEFHWBTNDYAS2Q","bare-skill-no-frontmatter","https:\u002F\u002Fgithub.com\u002FKemetra\u002FSeshat-BI\u002Fblob\u002Fmain\u002Ftests\u002Funit\u002Ffixtures\u002Fcapability_inventory\u002Forphan\u002F.claude\u002Fskills\u002Fbare-skill-no-frontmatter\u002FSKILL.md",[],[],{"skillNo":65,"name":48,"nameZh":6,"description":49,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":66,"categories":67,"keywords":68,"publishedAt":18},"01M77JW4TD1FNFSBNP5C0WXAA5","https:\u002F\u002Fgithub.com\u002FKemetra\u002FSeshat-BI\u002Fblob\u002Fmain\u002Ftests\u002Funit\u002Ffixtures\u002Fcapability_inventory\u002Fgood\u002F.claude\u002Fskills\u002Ffake-skill\u002FSKILL.md",[],[],{"skillNo":70,"name":71,"nameZh":6,"description":72,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":73,"categories":74,"keywords":75,"publishedAt":18},"01M77JVWCVE3WZ82ZNFCSFVDCP","retail-kpi-knowledge","Retail KPI \u002F metric-contract reasoning layer for BI agents in the Seshat BI project. Use when an agent must reason about the BUSINESS MEANING of a retail KPI before any implementation — defining a KPI in business terms, classifying additivity (fully\u002Fsemi\u002Fnon-additive), identifying grain, listing required source fields, resolving ambiguity (gross vs net, VAT, returns, cost method, same-store), choosing an MVP KPI pack, reviewing a metric contract for completeness, or preparing a clean handoff to an implementation layer. Also use it to classify KPI sufficiency or prepare an owner policy decision packet. This is a meaning\u002Fdefinition + review layer, not an executor: it does not write DAX\u002FSQL\u002FPython, grant readiness, approve policy, or design dashboards. See INDEX.md for live\u002Fplanned routes; registry.yaml is the authoritative machine-readable inventory.","https:\u002F\u002Fgithub.com\u002FKemetra\u002FSeshat-BI\u002Fblob\u002Fmain\u002Fskills\u002Fretail-kpi-knowledge\u002FSKILL.md",[],[],{"skillNo":77,"name":78,"nameZh":6,"description":79,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":80,"categories":81,"keywords":82,"publishedAt":18},"01M77JVNDM1NGYGSCPTBW9ZCF5","bi-sql-knowledge","SQL reasoning, validation, reconciliation, transformation, diagnostics, and anti-pattern knowledge layer for Seshat BI agents. Use when a task involves reasoning about source SQL: source profiling, table grain, keys & uniqueness, joins & fan-out amplification, aggregation correctness, COUNT\u002FNULL semantics, deduplication, validation queries, reconciliation queries, silver\u002Fgold transformation logic, window\u002Fdate-time analytics, SQL anti-patterns, or basic SQL performance reasoning, or interpretation of a supplied sanitized PostgreSQL JSON execution plan. The SQL counterpart to bi-dax-knowledge. It is a thinking & validation layer, not a database executor, runtime validator, dbt\u002FDagster project, or automatic tuning service.","https:\u002F\u002Fgithub.com\u002FKemetra\u002FSeshat-BI\u002Fblob\u002Fmain\u002Fskills\u002Fbi-sql-knowledge\u002FSKILL.md",[],[]]