privacy/workstreams/attribute-reporting.json

{"display_name":"Retailer Signals","external_boundaries":[{"content":"The package-supplied public image URL and ordinary network request metadata; downloaded image bytes remain local and are not sent back with the report","controls":["The image hydrator permits supported public raster URLs only and validates address, redirect, byte-size, and hash constraints.","Missing images remain a visible partial evidence state rather than triggering an undisclosed alternative route."],"destination":"Public product-image hosts named by the evidence package","id":"public-product-image-retrieval","kind":"public_research","optional":true,"purpose":"Download product images for local mapping evidence and report review","requires_confirmation":false,"runtime_profiles":["openai-codex","anthropic-cowork","openai-chatgpt","google-antigravity"]},{"content":"Published package requirements and ordinary package-index request metadata. Existing workflow arguments are forwarded to a local Python child; they are not included in the pip install command.","controls":["The direct CLI selects published core or registered component requirements before importing workflow modules.","The launcher preserves the working directory and arguments in the local child process; this does not redact arguments or change the workflow data boundary.","An existing matching runtime is reused. Setup installs into a fingerprinted user-scoped environment and propagates failure; it does not install into the case folder."],"destination":"Python Package Index (PyPI) or the index selected by the user's Python configuration","id":"direct-cli-python-dependency-setup","kind":"public_research","optional":false,"purpose":"Prepare the declared Python dependencies before running a documented workflow CLI","requires_confirmation":false,"runtime_profiles":["openai-codex","anthropic-cowork","openai-chatgpt","google-antigravity"]}],"governed_paths":["skills/attribute-reporting/SKILL.md","modules/attribute-reporting/skills/attribute-reporting/SKILL.md","modules/attribute-reporting/skills/attribute-reporting/references","modules/attribute-reporting/scripts","scripts/self_relaunch.py","scripts/managed_python_runtime.py","scripts/_managed_python_runtime.py","requirements.txt","components.json"],"model_context":{"classes":[{"content":"Retailer and category; structured product, variant, brand, description, price, promotion, ranking, cohort and taxonomy records; public image URLs and locally hydrated product images; package provenance and mapping history","id":"retailer-product-evidence","purpose":"Map product attributes and compare retailer cohorts","runtime_profiles":["openai-codex","anthropic-cowork","openai-chatgpt","google-antigravity"]},{"content":"Mapping tasks, decisions, evidence, local-image hashes, independent mapping reviews, computed tables, report model, claim ledger, caveats, selected examples, semantic review, browser QA, and correctness verdict","id":"mapping-and-report-judgement","purpose":"Resolve ambiguous taxonomy mappings and author and review a checked report","runtime_profiles":["openai-codex","anthropic-cowork","openai-chatgpt","google-antigravity"]}],"policy":"real_case_data_may_enter_selected_runtime_model_context"},"packaged_for_runtime":"openai-codex","public_explanations":{"clara-retailer-signals":{"copy":{"de":["Der Code berechnet Kohorten und Vergleiche; das Modell sieht Produktdaten und Nachweise für Zuordnung, Interpretation und Prüfung.","Das Modell kann Händler, Kategorie, Taxonomie, Produkt- und Markennamen, Beschreibungen, Merkmale, Kohortenzugehörigkeit, Quellzeilen, URLs und ausgewählte lokale Bilder erhalten. Es prüft jede ungelöste Zuordnungsaufgabe; eine zweite unabhängige Prüfung kontrolliert jede Entscheidung. Danach sieht es berechnete Vergleichstabellen, ausgewählte Aussagen, Bericht und Prüfmaterial.","Der Code berechnet Neuheiten gegenüber Restsortiment, Bestseller gegenüber anderen Produkten und verwandte Vergleiche über das gesamte Paket. Bilder und HTML-Berichte bleiben lokal und werden nicht zu Mparanza hochgeladen. Nur die vorgesehenen Zuordnungsartefakte dürfen die authentifizierte Dienstgrenze überschreiten; die Modellarbeit nutzt den ChatGPT-Plan des Nutzers, die Hilfsskripte rufen keine separate Modell-API auf."],"en":["Code calculates cohorts and comparisons; the model sees product records and evidence needed for mapping, interpretation, and review.","The model may receive retailer, category, taxonomy, product and brand names, descriptions, attributes, cohort membership, source rows, URLs, and selected local images. It reviews every unresolved mapping task, and a second independent review checks every decision; it then sees the calculated comparison tables, selected claims, report, and review material.","Code performs new-versus-rest, best-seller-versus-other, and related calculations across the complete package. Images and HTML reports remain local and are not uploaded to Mparanza. Only the defined mapping artifacts may cross the authenticated service boundary; model work uses the user's ChatGPT plan, and helper scripts make no separate model API call."],"es":["El código calcula cohortes y comparaciones; el modelo ve fichas de producto y evidencias necesarias para el mapeo, la interpretación y la revisión.","El modelo puede recibir minorista, categoría, taxonomía, nombres de productos y marcas, descripciones, atributos, pertenencia a cohortes, filas de origen, URL e imágenes locales seleccionadas. Examina cada tarea de mapeo no resuelta y una segunda revisión independiente comprueba cada decisión; después ve las tablas comparativas ya calculadas, las afirmaciones elegidas, el informe y el material de revisión.","El código realiza los cálculos de novedades frente al resto, superventas frente a otros productos y cálculos relacionados sobre todo el paquete. Las imágenes y los informes HTML permanecen locales y no se cargan en Mparanza. Solo los artefactos de mapeo previstos pueden cruzar el servicio autenticado; el trabajo del modelo usa el plan ChatGPT del usuario y los scripts no realizan una llamada API separada."],"fr":["Le code calcule cohortes et comparaisons ; le modèle voit les fiches produit et preuves nécessaires au mapping, à l'interprétation et à la revue.","Le modèle peut recevoir distributeur, catégorie, taxonomie, noms de produits et marques, descriptions, attributs, appartenance aux cohortes, lignes source, URL et images locales sélectionnées. Il examine chaque tâche de mapping non résolue, puis une seconde revue indépendante contrôle chaque décision ; il voit ensuite les tableaux comparatifs déjà calculés, les affirmations choisies, le rapport et les éléments de revue.","Le code effectue les calculs nouveautés-reste, meilleures ventes-autres produits et les calculs associés sur tout le paquet. Les images et rapports HTML restent locaux et ne sont pas envoyés à Mparanza. Seuls les artefacts de mapping prévus peuvent franchir le service authentifié ; le travail du modèle utilise le forfait ChatGPT de l'utilisateur et les scripts n'effectuent aucun appel API séparé."],"it":["Il codice calcola coorti e confronti; il modello vede le schede prodotto e le evidenze necessarie per mapping, interpretazione e revisione.","Possono arrivare al modello retailer, categoria, tassonomia, nomi e brand dei prodotti, descrizioni, attributi, appartenenza alle coorti, righe sorgente, URL e immagini locali selezionate. Il modello esamina ogni attività di mapping non risolta e una seconda revisione indipendente controlla ogni decisione; vede poi le tabelle di confronto già calcolate, le affermazioni scelte, il report e i materiali di revisione.","I calcoli su novità, resto dell'assortimento, best seller e altri prodotti sono eseguiti dal codice sull'intero pacchetto. Immagini e report HTML restano locali e non vengono caricati su Mparanza. Solo gli artefatti di mapping previsti possono attraversare il servizio autenticato; il lavoro del modello usa il piano ChatGPT dell'utente e gli script non effettuano una chiamata API separata."]},"status":"relevant"}},"review":{"basis":"external_boundary_review_of_workflow_source","reviewed_at":"2026-10-29","reviewed_by":"privacy-surface-review","source_fingerprint":"849b64a08dbbf53a4c4330e3a8ef20d26a690399b3f7a0427a489e4f6632afe0"},"role":"workflow","runtime_profiles":["openai-codex","anthropic-cowork","openai-chatgpt","google-antigravity"],"schema_version":3,"security_controls":[{"control":"Authentication cookies are kept in a private local auth directory and attached only to the exact approved Mparanza origin.","id":"origin-bound-auth"},{"control":"Downloads are size-limited, checksum-verified, safely extracted, and bound to local transport and package provenance.","id":"checksummed-packages"}],"service_ids":["managed-python-runtime","operational-incidents","retail-data","run-receipt-stamping","workflow-usage"],"source_contract_sha256":"259c2e4cea1d8593e8dfc6c0f801557eeb6583c1766912a6c65e815d78306d4c","workstream":"attribute-reporting"}