{
  "canonical_name": "ollama/ollama",
  "compilation_id": "pack_d52552baf5c24580bd20cfc117b922d5",
  "created_at": "2026-05-11T12:57:37.050135+00:00",
  "created_by": "project-pack-compiler",
  "feedback": {
    "carrier_selection_notes": [
      "viable_asset_types=prompt, recipe, host_instruction, eval, preflight",
      "recommended_asset_types=prompt, recipe, host_instruction, eval, preflight"
    ],
    "evidence_delta": {
      "confirmed_claims": [
        "identity_anchor_present",
        "capability_and_host_targets_present",
        "install_path_declared_or_better"
      ],
      "missing_required_fields": [],
      "must_verify_forwarded": [
        "Run or inspect `pip install ollama` in an isolated environment.",
        "Confirm the project exposes the claimed capability to at least one target host."
      ],
      "quickstart_execution_scope": "allowlisted_sandbox_smoke",
      "sandbox_command": "pip install ollama",
      "sandbox_container_image": "python:3.12-slim",
      "sandbox_execution_backend": "docker",
      "sandbox_planner_decision": "deterministic_isolated_install",
      "sandbox_validation_id": "sbx_53b8bb99a8db4523ba9fe5cc31fb91ed"
    },
    "feedback_event_type": "project_pack_compilation_feedback",
    "learning_candidate_reasons": [],
    "template_gaps": []
  },
  "identity": {
    "canonical_id": "project_4c2c0148df16816a6d9dfb5e2276a5f1",
    "canonical_name": "ollama/ollama",
    "homepage_url": null,
    "license": "unknown",
    "repo_url": "https://github.com/ollama/ollama",
    "slug": "ollama",
    "source_packet_id": "phit_fbe2c9c27a834e12906c0f8e9a0f4221",
    "source_validation_id": "dval_79f405cb6efb4ba58f5cb60a06b5c482"
  },
  "merchandising": {
    "best_for": "需要软件开发与交付能力，并使用 local_cli的用户",
    "github_forks": 16091,
    "github_stars": 171296,
    "one_liner_en": "Ollama capability pack for local model serving, CLI/API checks, model pull/run workflows, resource budgets, and rollback boundaries.",
    "one_liner_zh": "Get up and running with Kimi-K2.5, GLM-5, MiniMax, DeepSeek, gpt-oss, Qwen, Gemma and other models.",
    "primary_category": {
      "category_id": "software-development",
      "confidence": "medium",
      "name_en": "Software Development",
      "name_zh": "软件开发与交付",
      "reason": "matched_keywords:git, cli"
    },
    "target_user": "使用 local_cli 等宿主 AI 的用户",
    "title_en": "Ollama Local Model Runtime Pack",
    "title_zh": "ollama 能力包",
    "visible_tags": [
      "Ollama",
      "Local models",
      "Model serving",
      "CLI",
      "API compatibility",
      "Resource budget"
    ],
    "subtitle_en": "Ollama capability pack for local model serving, CLI/API checks, model pull/run workflows, resource budgets, and rollback boundaries.",
    "best_for_en": "Developers who need local model serving, model management, or API-compatible experiments with explicit resource and port boundaries.",
    "capability_type_en": "Local model runtime setup, model pull/run checks, CLI/API verification, resource budget review, and rollback guidance"
  },
  "packet_id": "phit_fbe2c9c27a834e12906c0f8e9a0f4221",
  "page_model": {
    "artifacts": {
      "artifact_slug": "ollama",
      "files": [
        "PROJECT_PACK.json",
        "QUICK_START.md",
        "PROMPT_PREVIEW.md",
        "HUMAN_MANUAL.md",
        "AI_CONTEXT_PACK.md",
        "BOUNDARY_RISK_CARD.md",
        "PITFALL_LOG.md",
        "REPO_INSPECTION.json",
        "REPO_INSPECTION.md",
        "CAPABILITY_CONTRACT.json",
        "EVIDENCE_INDEX.json",
        "CLAIM_GRAPH.json"
      ],
      "required_files": [
        "PROJECT_PACK.json",
        "QUICK_START.md",
        "PROMPT_PREVIEW.md",
        "HUMAN_MANUAL.md",
        "AI_CONTEXT_PACK.md",
        "BOUNDARY_RISK_CARD.md",
        "PITFALL_LOG.md",
        "REPO_INSPECTION.json"
      ]
    },
    "detail": {
      "capability_source": "Project Hit Packet + DownstreamValidationResult",
      "commands": [
        {
          "command": "pip install ollama",
          "label": "Python / pip · 官方安装入口",
          "source": "https://github.com/ollama/ollama#readme",
          "verified": true
        }
      ],
      "display_tags": [
        "Ollama",
        "Local models",
        "Model serving",
        "CLI",
        "API compatibility",
        "Resource budget"
      ],
      "eyebrow": "软件开发与交付",
      "glance": [
        {
          "body": "判断自己是不是目标用户。",
          "label": "最适合谁",
          "value": "需要软件开发与交付能力，并使用 local_cli的用户"
        },
        {
          "body": "先理解能力边界，再决定是否继续。",
          "label": "核心价值",
          "value": "Get up and running with Kimi-K2.5, GLM-5, MiniMax, DeepSeek, gpt-oss, Qwen, Gemma and other models."
        },
        {
          "body": "未完成验证前保持审慎。",
          "label": "继续前",
          "value": "publish to Doramagic.ai project surfaces"
        }
      ],
      "guardrail_source": "Boundary & Risk Card",
      "guardrails": [
        {
          "title": "Keep source authority visible",
          "body": "Use Ollama Local Model Runtime Pack with the upstream repository as the final source of truth."
        },
        {
          "title": "Start with a reversible check",
          "body": "Pull and run one small model on an isolated port, verify health/API behavior, then stop and clean up the runtime."
        },
        {
          "title": "Do not overstate evidence",
          "body": "The main risk is unbounded CPU/GPU, memory, port exposure, model storage, or API compatibility assumptions."
        }
      ],
      "mode": "Local model runtime setup, model pull/run checks, CLI/API verification, resource budget review, and rollback guidance",
      "pitfall_log": {
        "items": [
          {
            "severity": "medium",
            "category": "Verification boundary",
            "title": "Do not skip the first safe check",
            "body": "Pull and run one small model on an isolated port, verify health/API behavior, then stop and clean up the runtime.",
            "user_impact": "The main risk is unbounded CPU/GPU, memory, port exposure, model storage, or API compatibility assumptions.",
            "suggested_check": "Run the smallest reversible fixture before real data."
          },
          {
            "severity": "medium",
            "category": "Source authority",
            "title": "Use upstream as final truth",
            "body": "Generated canary copy is a search and AI retrieval contract, not a replacement for upstream docs.",
            "user_impact": "Users may follow stale commands if source authority is hidden.",
            "suggested_check": "Open the upstream repository before running commands."
          },
          {
            "severity": "medium",
            "category": "Rollback",
            "title": "Define cleanup before execution",
            "body": "Every first run needs a timeout, cleanup path, and output boundary.",
            "user_impact": "Generated files or runtime state can linger after a failed trial.",
            "suggested_check": "Write the cleanup step next to the command."
          },
          {
            "severity": "low",
            "category": "Evidence gap",
            "title": "Missing evidence is not a positive signal",
            "body": "The page must expose missing evidence rather than turning it into a recommendation.",
            "user_impact": "Users may overtrust a generated capability pack.",
            "suggested_check": "List missing evidence before go/no-go."
          },
          {
            "severity": "low",
            "category": "Semantic identity",
            "title": "Keep the project in its true category",
            "body": "This page must describe Local model runtime setup, model pull/run checks, CLI/API verification, resource budget review, and rollback guidance, not an unrelated automation category.",
            "user_impact": "Search and AI retrieval can route users to the wrong use case.",
            "suggested_check": "Compare title, tags, and schema against the semantic profile."
          }
        ]
      },
      "snapshot": {
        "contributors": 600,
        "forks": 16091,
        "license": "unknown",
        "note": "GitHub API 快照，非实时质量证明；用于开工前背景判断。",
        "stars": 171296,
        "open_issues": 3238,
        "pushed_at": null
      },
      "source_url": "https://github.com/ollama/ollama",
      "steps": [
        {
          "title": "Runtime setup",
          "body": "Start one small model on an isolated port and record the exact pull/run commands.",
          "code": "runtime-setup"
        },
        {
          "title": "API and CLI check",
          "body": "Verify CLI behavior, health/API compatibility, model storage path, and shutdown behavior.",
          "code": "api-cli-check"
        },
        {
          "title": "Resource budget",
          "body": "Measure CPU/GPU, memory, concurrency, and rollback before using larger models or shared ports.",
          "code": "resource-budget"
        }
      ],
      "subtitle": "Ollama capability pack for local model serving, CLI/API checks, model pull/run workflows, resource budgets, and rollback boundaries.",
      "title": "Ollama Local Model Runtime Pack",
      "trial_prompt": "Prompt preview for Ollama Local Model Runtime Pack\n\nGoal: evaluate whether Ollama Local Model Runtime Pack from ollama/ollama should be used for my current task before taking action.\n\nYou are using a Doramagic English canary Project Pack. Before taking action, separate upstream facts from Doramagic interpretation. Use the source repository https://github.com/ollama/ollama as the evidence anchor. Check the capability boundary: Local model runtime setup, model pull/run checks, CLI/API verification, resource budget review, and rollback guidance. Start with this safe step: Pull and run one small model on an isolated port, verify health/API behavior, then stop and clean up the runtime.. Identify missing evidence, state the main risk, and propose one tiny reversible verification fixture. Do not run install commands, write files, use credentials, access a network service, or modify persistent state unless I explicitly approve the command and rollback plan.\n\nReturn: a practical go/no-go review, the first verification command or manual check, expected output, rollback path, and the evidence still missing.",
      "voices": [
        {
          "body": "来源平台：github。github/github_issue: Registering fine-tuned models（https://github.com/ollama/ollama/issues/16095）；github/github_issue: Ollama Cloud: Frequent 503 errors making cloud models unreliable（https://github.com/ollama/ollama/issues/15419）；github/github_issue: Install libs only for detected arch.（https://github.com/ollama/ollama/issues/16098）；github/github_issue: Not compatible with Glaude code Cli when using local model（https://github.com/ollama/ollama/issues/16094）；github/github_issue: ollama launch claude: fails with API Error 400 when user has CLAUDE_CODE（https://github.com/ollama/ollama/issues/16097）；github/github_issue: 0.23.1 : mlx runner failed（https://github.com/ollama/ollama/issues/16007）；github/github_issue: Featured your project on osalt.dev — README badge available if you'd lik（https://github.com/ollama/ollama/issues/16092）；github/github_issue: mistral-medium-3.5 - Produces nonsense outputs（https://github.com/ollama/ollama/issues/15975）；github/github_issue: Support `ppc64le` architecture（https://github.com/ollama/ollama/issues/796）；github/github_issue: Running qwen3.6:27b-q8_0 produces also gibberish on an AMD Ryzen AI Max+（https://github.com/ollama/ollama/issues/15903）；github/github_issue: Running qwen3.6:27b-bf16 on an AMD Ryzen AI Max leads to gibberish（https://github.com/ollama/ollama/issues/15879）；github/github_issue: SIGSEGV in MLX VAE decode after diffusion steps complete on M4 Pro (macO（https://github.com/ollama/ollama/issues/16093）。这些是项目级外部声音，不作为单独质量证明。",
          "items": [
            {
              "kind": "github_issue",
              "source": "github",
              "title": "Registering fine-tuned models",
              "url": "https://github.com/ollama/ollama/issues/16095"
            },
            {
              "kind": "github_issue",
              "source": "github",
              "title": "Ollama Cloud: Frequent 503 errors making cloud models unreliable",
              "url": "https://github.com/ollama/ollama/issues/15419"
            },
            {
              "kind": "github_issue",
              "source": "github",
              "title": "Install libs only for detected arch.",
              "url": "https://github.com/ollama/ollama/issues/16098"
            },
            {
              "kind": "github_issue",
              "source": "github",
              "title": "Not compatible with Glaude code Cli when using local model",
              "url": "https://github.com/ollama/ollama/issues/16094"
            },
            {
              "kind": "github_issue",
              "source": "github",
              "title": "ollama launch claude: fails with API Error 400 when user has CLAUDE_CODE",
              "url": "https://github.com/ollama/ollama/issues/16097"
            },
            {
              "kind": "github_issue",
              "source": "github",
              "title": "0.23.1 : mlx runner failed",
              "url": "https://github.com/ollama/ollama/issues/16007"
            },
            {
              "kind": "github_issue",
              "source": "github",
              "title": "Featured your project on osalt.dev — README badge available if you'd lik",
              "url": "https://github.com/ollama/ollama/issues/16092"
            },
            {
              "kind": "github_issue",
              "source": "github",
              "title": "mistral-medium-3.5 - Produces nonsense outputs",
              "url": "https://github.com/ollama/ollama/issues/15975"
            },
            {
              "kind": "github_issue",
              "source": "github",
              "title": "Support `ppc64le` architecture",
              "url": "https://github.com/ollama/ollama/issues/796"
            },
            {
              "kind": "github_issue",
              "source": "github",
              "title": "Running qwen3.6:27b-q8_0 produces also gibberish on an AMD Ryzen AI Max+",
              "url": "https://github.com/ollama/ollama/issues/15903"
            },
            {
              "kind": "github_issue",
              "source": "github",
              "title": "Running qwen3.6:27b-bf16 on an AMD Ryzen AI Max leads to gibberish",
              "url": "https://github.com/ollama/ollama/issues/15879"
            },
            {
              "kind": "github_issue",
              "source": "github",
              "title": "SIGSEGV in MLX VAE decode after diffusion steps complete on M4 Pro (macO",
              "url": "https://github.com/ollama/ollama/issues/16093"
            }
          ],
          "status": "已收录 12 条来源",
          "title": "社区讨论"
        }
      ],
      "title_en": "Ollama Local Model Runtime Pack",
      "subtitle_en": "Ollama capability pack for local model serving, CLI/API checks, model pull/run workflows, resource budgets, and rollback boundaries.",
      "trial_prompt_en": "Prompt preview for Ollama Local Model Runtime Pack\n\nGoal: evaluate whether Ollama Local Model Runtime Pack from ollama/ollama should be used for my current task before taking action.\n\nYou are using a Doramagic English canary Project Pack. Before taking action, separate upstream facts from Doramagic interpretation. Use the source repository https://github.com/ollama/ollama as the evidence anchor. Check the capability boundary: Local model runtime setup, model pull/run checks, CLI/API verification, resource budget review, and rollback guidance. Start with this safe step: Pull and run one small model on an isolated port, verify health/API behavior, then stop and clean up the runtime.. Identify missing evidence, state the main risk, and propose one tiny reversible verification fixture. Do not run install commands, write files, use credentials, access a network service, or modify persistent state unless I explicitly approve the command and rollback plan.\n\nReturn: a practical go/no-go review, the first verification command or manual check, expected output, rollback path, and the evidence still missing."
    },
    "homepage_card": {
      "category": "软件开发与交付",
      "desc": "Get up and running with Kimi-K2.5, GLM-5, MiniMax, DeepSeek, gpt-oss, Qwen, Gemma and other models.",
      "effort": "安装已验证",
      "forks": 16063,
      "icon": "code",
      "name": "ollama 能力包",
      "risk": "可发布",
      "slug": "ollama",
      "stars": 171181,
      "tags": [
        "浏览器 Agent",
        "网页任务自动化",
        "浏览器自动化",
        "断点恢复流程",
        "评测体系"
      ],
      "thumb": "gray",
      "type": "Prompt Preview"
    },
    "manual": {
      "markdown": "# Ollama Local Model Runtime Pack Human Manual\n\nGenerated for Doramagic SEO/GEO English canary validation from the existing Project Pack, semantic profile, quality gate, and source repository reference.\n\n## Table of Contents\n\n- [Project identity](#project-identity)\n- [Capability boundary](#capability-boundary)\n- [Evidence and source policy](#evidence-and-source-policy)\n- [Pre-install verification path](#pre-install-verification-path)\n- [AI host handoff](#ai-host-handoff)\n- [Doramagic Pitfall Log](#doramagic-pitfall-log)\n- [Acceptance checklist](#acceptance-checklist)\n\n## Project identity\n\nProject: Ollama Local Model Runtime Pack\n\nCanonical repository: ollama/ollama\n\nSource URL: https://github.com/ollama/ollama\n\nWhat it is: Ollama is a local model runtime for downloading, running, and serving language models through CLI and API workflows.\n\nBest fit: Developers who need local model serving, model management, or API-compatible experiments with explicit resource and port boundaries.\n\nNot for: Not for workflows that only need a hosted model API, or environments that cannot maintain a local runtime, model storage, and service port.\n\nThe English canary page exists to make the project identity explicit for search engines and AI retrieval systems. It should preserve the upstream repository link, visible source evidence, and user-facing verification boundary. It must not imply that Doramagic has completed a fresh production deployment, live benchmark, or local installation beyond the evidence already carried by the source Project Pack.\n\n## Capability boundary\n\nCapability added to an AI workflow: Local model runtime setup, model pull/run checks, CLI/API verification, resource budget review, and rollback guidance\n\nPrimary risk: The main risk is unbounded CPU/GPU, memory, port exposure, model storage, or API compatibility assumptions.\n\nSemantic tags: Ollama, Local models, Model serving, CLI, API compatibility, Resource budget\n\n1. Runtime setup: Start one small model on an isolated port and record the exact pull/run commands.\n2. API and CLI check: Verify CLI behavior, health/API compatibility, model storage path, and shutdown behavior.\n3. Resource budget: Measure CPU/GPU, memory, concurrency, and rollback before using larger models or shared ports.\n\nThe boundary is deliberately narrow. A user should be able to decide whether the project is relevant, copy a prompt into an AI host, read the manual, and verify one small task before installing anything in a primary environment. This is not a guarantee that the upstream project is safe for every workload.\n\n## Evidence and source policy\n\nDoramagic uses the existing Project Pack as the evidence envelope for this English canary. The generated page keeps the upstream repository visible, keeps the canonical name stable, and uses the semantic profile only to prevent known identity contamination such as browser-automation copy on non-browser projects.\n\nSource-backed fields used here include identity, repository URL, quality gate status, commands when available, guardrails, pitfall items, and the semantic canary profile. When a command or risk item is missing, the page must disclose that absence and route the user to sandbox verification instead of inventing a happy path.\n\n## Pre-install verification path\n\nFirst safe step: Pull and run one small model on an isolated port, verify health/API behavior, then stop and clean up the runtime.\n\n- Check 1: Python / pip · 官方安装入口: `pip install ollama` (source: https://github.com/ollama/ollama#readme).\n\nBefore using real data, run the smallest reversible check possible. Keep secrets out of the first run, record the exact command or API call, record expected output, define a timeout, and decide how to clean up generated files or runtime state. If the upstream quick start changes, the source repository should override this generated canary text.\n\n## AI host handoff\n\nUse this pack as portable context, not as an automatic install instruction. A safe AI-host handoff should include the source URL, the capability boundary, the first safe step, known risks, and an explicit instruction to ask before running commands that touch credentials, files, network, or persistent state.\n\nPrompt preview users should ask the host AI to produce a go/no-go decision, list missing evidence, identify a tiny verification fixture, and separate upstream facts from Doramagic interpretation. This keeps the page useful for ChatGPT, Claude, Gemini, Codex, Cursor, and other hosts without locking the asset to one provider.\n\n## Doramagic Pitfall Log\n\n- Pitfall 1: Do not skip the first safe check Pull and run one small model on an isolated port, verify health/API behavior, then stop and clean up the runtime. The main risk is unbounded CPU/GPU, memory, port exposure, model storage, or API compatibility assumptions. Run the smallest reversible fixture before real data.\n- Pitfall 2: Use upstream as final truth Generated canary copy is a search and AI retrieval contract, not a replacement for upstream docs. Users may follow stale commands if source authority is hidden. Open the upstream repository before running commands.\n- Pitfall 3: Define cleanup before execution Every first run needs a timeout, cleanup path, and output boundary. Generated files or runtime state can linger after a failed trial. Write the cleanup step next to the command.\n- Pitfall 4: Missing evidence is not a positive signal The page must expose missing evidence rather than turning it into a recommendation. Users may overtrust a generated capability pack. List missing evidence before go/no-go.\n- Pitfall 5: Keep the project in its true category This page must describe Local model runtime setup, model pull/run checks, CLI/API verification, resource budget review, and rollback guidance, not an unrelated automation category. Search and AI retrieval can route users to the wrong use case. Compare title, tags, and schema against the semantic profile.\n\nGuardrails:\n\n- Use Ollama Local Model Runtime Pack with the upstream repository as the final source of truth.\n- Pull and run one small model on an isolated port, verify health/API behavior, then stop and clean up the runtime.\n- The main risk is unbounded CPU/GPU, memory, port exposure, model storage, or API compatibility assumptions.\n\nThe pitfall log is intentionally conservative. It converts missing evidence and boundary uncertainty into checks the user can run. It should not be rewritten into first-person testing claims unless a fresh sandbox run, trace, and artifact manifest prove that claim.\n\n## Acceptance checklist\n\n- The page title follows the Doramagic.ai title format.\n- The page exposes SoftwareSourceCode, TechArticle, BreadcrumbList, and FAQPage structured data.\n- The page links back to https://github.com/ollama/ollama.\n- The page has a Markdown alternate route for AI consumers.\n- The page keeps Ollama Local Model Runtime Pack associated with its true semantic identity: Local model runtime setup, model pull/run checks, CLI/API verification, resource budget review, and rollback guidance.\n- The page avoids forbidden identity drift such as browser automation language when the source project is not a browser automation project.\n- The page remains reversible: remove the generated English pack root and the build falls back to the original source-cache state.\n",
      "markdown_key": "ollama",
      "pages": "draft",
      "source_refs": [
        {
          "label": "ollama/ollama",
          "url": "https://github.com/ollama/ollama"
        }
      ],
      "summary": "Ollama Local Model Runtime Pack source-backed English canary manual for SEO/GEO validation.",
      "title": "Ollama Local Model Runtime Pack Human Manual",
      "toc": [
        "Table of Contents",
        "Project identity",
        "Capability boundary",
        "Evidence and source policy",
        "Pre-install verification path",
        "AI host handoff",
        "Doramagic Pitfall Log",
        "Acceptance checklist"
      ]
    }
  },
  "quality_gate": {
    "blocking_gaps": [],
    "category_confidence": "medium",
    "compile_status": "ready_for_review",
    "five_assets_present": true,
    "install_sandbox_verified": true,
    "missing_evidence": [],
    "next_action": "publish English canary page after SEO/GEO acceptance passes",
    "prompt_preview_boundary_ok": true,
    "publish_status": "publishable",
    "quick_start_verified": true,
    "repo_clone_verified": true,
    "repo_commit": "c2f2d90a679317e443b0c35aa6d6d747cdd6c57b",
    "repo_inspection_error": null,
    "repo_inspection_files": [
      "Dockerfile",
      "README.md",
      "docs/import.mdx",
      "docs/template.mdx",
      "docs/macos.mdx",
      "docs/cloud.mdx",
      "docs/troubleshooting.mdx",
      "docs/gpu.mdx",
      "docs/docker.mdx",
      "docs/api.md",
      "docs/cli.mdx",
      "docs/windows.mdx",
      "docs/examples.md",
      "docs/index.mdx",
      "docs/README.md",
      "docs/development.md",
      "docs/docs.json",
      "docs/quickstart.mdx",
      "docs/openapi.yaml",
      "docs/context-length.mdx",
      "docs/modelfile.mdx",
      "docs/linux.mdx",
      "docs/faq.mdx",
      "docs/api/openai-compatibility.mdx",
      "docs/api/errors.mdx",
      "docs/api/authentication.mdx",
      "docs/api/anthropic-compatibility.mdx",
      "docs/api/introduction.mdx",
      "docs/api/usage.mdx",
      "docs/api/streaming.mdx",
      "docs/capabilities/structured-outputs.mdx",
      "docs/capabilities/thinking.mdx",
      "docs/capabilities/web-search.mdx",
      "docs/capabilities/embeddings.mdx",
      "docs/capabilities/tool-calling.mdx",
      "docs/capabilities/vision.mdx",
      "docs/capabilities/streaming.mdx",
      "docs/integrations/opencode.mdx",
      "docs/integrations/nemoclaw.mdx",
      "docs/integrations/onyx.mdx"
    ],
    "repo_inspection_verified": true,
    "review_reasons": [],
    "tag_count_ok": true,
    "unsupported_claims": []
  },
  "schema_version": "0.1",
  "user_assets": {
    "ai_context_pack": {
      "filename": "AI_CONTEXT_PACK.md",
      "title": "Ollama Local Model Runtime Pack AI Context Pack",
      "summary": "Portable English context for AI hosts.",
      "markdown": "# Ollama Local Model Runtime Pack AI Context Pack\n\n- Canonical project: ollama/ollama\n- Source URL: https://github.com/ollama/ollama\n- What it is: Ollama is a local model runtime for downloading, running, and serving language models through CLI and API workflows.\n- Best fit: Developers who need local model serving, model management, or API-compatible experiments with explicit resource and port boundaries.\n- Capability: Local model runtime setup, model pull/run checks, CLI/API verification, resource budget review, and rollback guidance\n- First safe step: Pull and run one small model on an isolated port, verify health/API behavior, then stop and clean up the runtime.\n- Top risk: The main risk is unbounded CPU/GPU, memory, port exposure, model storage, or API compatibility assumptions.\n\nUse this context in an AI host to evaluate the project before installation. Ask for evidence gaps, a tiny verification fixture, a rollback path, and a clear distinction between upstream facts and Doramagic interpretation.\n"
    },
    "boundary_risk_card": {
      "filename": "BOUNDARY_RISK_CARD.md",
      "title": "Ollama Local Model Runtime Pack Boundary and Risk Card",
      "summary": "English risk boundary extracted from the semantic profile and existing pack guardrails.",
      "markdown": "# Ollama Local Model Runtime Pack Boundary and Risk Card\n\nSource: https://github.com/ollama/ollama\nCanonical project: ollama/ollama\n\nMain capability: Local model runtime setup, model pull/run checks, CLI/API verification, resource budget review, and rollback guidance\n\nMain risk: The main risk is unbounded CPU/GPU, memory, port exposure, model storage, or API compatibility assumptions.\n\nFirst safe step: Pull and run one small model on an isolated port, verify health/API behavior, then stop and clean up the runtime.\n\n## Source guardrails\n\n- Use Ollama Local Model Runtime Pack with the upstream repository as the final source of truth.\n- Pull and run one small model on an isolated port, verify health/API behavior, then stop and clean up the runtime.\n- The main risk is unbounded CPU/GPU, memory, port exposure, model storage, or API compatibility assumptions.\n"
    },
    "human_manual": {
      "filename": "HUMAN_MANUAL.md",
      "title": "Ollama Local Model Runtime Pack Human Manual",
      "summary": "Source-backed English canary manual generated from the existing Project Pack, semantic profile, and public repository reference.",
      "markdown": "# Ollama Local Model Runtime Pack Human Manual\n\nGenerated for Doramagic SEO/GEO English canary validation from the existing Project Pack, semantic profile, quality gate, and source repository reference.\n\n## Table of Contents\n\n- [Project identity](#project-identity)\n- [Capability boundary](#capability-boundary)\n- [Evidence and source policy](#evidence-and-source-policy)\n- [Pre-install verification path](#pre-install-verification-path)\n- [AI host handoff](#ai-host-handoff)\n- [Doramagic Pitfall Log](#doramagic-pitfall-log)\n- [Acceptance checklist](#acceptance-checklist)\n\n## Project identity\n\nProject: Ollama Local Model Runtime Pack\n\nCanonical repository: ollama/ollama\n\nSource URL: https://github.com/ollama/ollama\n\nWhat it is: Ollama is a local model runtime for downloading, running, and serving language models through CLI and API workflows.\n\nBest fit: Developers who need local model serving, model management, or API-compatible experiments with explicit resource and port boundaries.\n\nNot for: Not for workflows that only need a hosted model API, or environments that cannot maintain a local runtime, model storage, and service port.\n\nThe English canary page exists to make the project identity explicit for search engines and AI retrieval systems. It should preserve the upstream repository link, visible source evidence, and user-facing verification boundary. It must not imply that Doramagic has completed a fresh production deployment, live benchmark, or local installation beyond the evidence already carried by the source Project Pack.\n\n## Capability boundary\n\nCapability added to an AI workflow: Local model runtime setup, model pull/run checks, CLI/API verification, resource budget review, and rollback guidance\n\nPrimary risk: The main risk is unbounded CPU/GPU, memory, port exposure, model storage, or API compatibility assumptions.\n\nSemantic tags: Ollama, Local models, Model serving, CLI, API compatibility, Resource budget\n\n1. Runtime setup: Start one small model on an isolated port and record the exact pull/run commands.\n2. API and CLI check: Verify CLI behavior, health/API compatibility, model storage path, and shutdown behavior.\n3. Resource budget: Measure CPU/GPU, memory, concurrency, and rollback before using larger models or shared ports.\n\nThe boundary is deliberately narrow. A user should be able to decide whether the project is relevant, copy a prompt into an AI host, read the manual, and verify one small task before installing anything in a primary environment. This is not a guarantee that the upstream project is safe for every workload.\n\n## Evidence and source policy\n\nDoramagic uses the existing Project Pack as the evidence envelope for this English canary. The generated page keeps the upstream repository visible, keeps the canonical name stable, and uses the semantic profile only to prevent known identity contamination such as browser-automation copy on non-browser projects.\n\nSource-backed fields used here include identity, repository URL, quality gate status, commands when available, guardrails, pitfall items, and the semantic canary profile. When a command or risk item is missing, the page must disclose that absence and route the user to sandbox verification instead of inventing a happy path.\n\n## Pre-install verification path\n\nFirst safe step: Pull and run one small model on an isolated port, verify health/API behavior, then stop and clean up the runtime.\n\n- Check 1: Python / pip · 官方安装入口: `pip install ollama` (source: https://github.com/ollama/ollama#readme).\n\nBefore using real data, run the smallest reversible check possible. Keep secrets out of the first run, record the exact command or API call, record expected output, define a timeout, and decide how to clean up generated files or runtime state. If the upstream quick start changes, the source repository should override this generated canary text.\n\n## AI host handoff\n\nUse this pack as portable context, not as an automatic install instruction. A safe AI-host handoff should include the source URL, the capability boundary, the first safe step, known risks, and an explicit instruction to ask before running commands that touch credentials, files, network, or persistent state.\n\nPrompt preview users should ask the host AI to produce a go/no-go decision, list missing evidence, identify a tiny verification fixture, and separate upstream facts from Doramagic interpretation. This keeps the page useful for ChatGPT, Claude, Gemini, Codex, Cursor, and other hosts without locking the asset to one provider.\n\n## Doramagic Pitfall Log\n\n- Pitfall 1: Do not skip the first safe check Pull and run one small model on an isolated port, verify health/API behavior, then stop and clean up the runtime. The main risk is unbounded CPU/GPU, memory, port exposure, model storage, or API compatibility assumptions. Run the smallest reversible fixture before real data.\n- Pitfall 2: Use upstream as final truth Generated canary copy is a search and AI retrieval contract, not a replacement for upstream docs. Users may follow stale commands if source authority is hidden. Open the upstream repository before running commands.\n- Pitfall 3: Define cleanup before execution Every first run needs a timeout, cleanup path, and output boundary. Generated files or runtime state can linger after a failed trial. Write the cleanup step next to the command.\n- Pitfall 4: Missing evidence is not a positive signal The page must expose missing evidence rather than turning it into a recommendation. Users may overtrust a generated capability pack. List missing evidence before go/no-go.\n- Pitfall 5: Keep the project in its true category This page must describe Local model runtime setup, model pull/run checks, CLI/API verification, resource budget review, and rollback guidance, not an unrelated automation category. Search and AI retrieval can route users to the wrong use case. Compare title, tags, and schema against the semantic profile.\n\nGuardrails:\n\n- Use Ollama Local Model Runtime Pack with the upstream repository as the final source of truth.\n- Pull and run one small model on an isolated port, verify health/API behavior, then stop and clean up the runtime.\n- The main risk is unbounded CPU/GPU, memory, port exposure, model storage, or API compatibility assumptions.\n\nThe pitfall log is intentionally conservative. It converts missing evidence and boundary uncertainty into checks the user can run. It should not be rewritten into first-person testing claims unless a fresh sandbox run, trace, and artifact manifest prove that claim.\n\n## Acceptance checklist\n\n- The page title follows the Doramagic.ai title format.\n- The page exposes SoftwareSourceCode, TechArticle, BreadcrumbList, and FAQPage structured data.\n- The page links back to https://github.com/ollama/ollama.\n- The page has a Markdown alternate route for AI consumers.\n- The page keeps Ollama Local Model Runtime Pack associated with its true semantic identity: Local model runtime setup, model pull/run checks, CLI/API verification, resource budget review, and rollback guidance.\n- The page avoids forbidden identity drift such as browser automation language when the source project is not a browser automation project.\n- The page remains reversible: remove the generated English pack root and the build falls back to the original source-cache state.\n"
    },
    "pitfall_log": {
      "asset_id": "pitfall_log",
      "filename": "PITFALL_LOG.md",
      "markdown": "",
      "summary": "用户实践前最可能遇到的身份、安装、配置、运行和安全坑。",
      "title": "Pitfall Log / 踩坑日志"
    },
    "prompt_preview": {
      "filename": "PROMPT_PREVIEW.md",
      "title": "Ollama Local Model Runtime Pack Prompt Preview",
      "summary": "English prompt preview for pre-install evaluation.",
      "markdown": "Prompt preview for Ollama Local Model Runtime Pack\n\nGoal: evaluate whether Ollama Local Model Runtime Pack from ollama/ollama should be used for my current task before taking action.\n\nYou are using a Doramagic English canary Project Pack. Before taking action, separate upstream facts from Doramagic interpretation. Use the source repository https://github.com/ollama/ollama as the evidence anchor. Check the capability boundary: Local model runtime setup, model pull/run checks, CLI/API verification, resource budget review, and rollback guidance. Start with this safe step: Pull and run one small model on an isolated port, verify health/API behavior, then stop and clean up the runtime.. Identify missing evidence, state the main risk, and propose one tiny reversible verification fixture. Do not run install commands, write files, use credentials, access a network service, or modify persistent state unless I explicitly approve the command and rollback plan.\n\nReturn: a practical go/no-go review, the first verification command or manual check, expected output, rollback path, and the evidence still missing."
    },
    "quick_start": {
      "filename": "QUICK_START.md",
      "title": "Ollama Local Model Runtime Pack Quick Start",
      "summary": "Source-linked pre-install check sequence.",
      "markdown": "# Ollama Local Model Runtime Pack Quick Start\n\nRepository: https://github.com/ollama/ollama\n\n1. Confirm that ollama/ollama is the intended upstream project.\n2. Read the capability boundary: Local model runtime setup, model pull/run checks, CLI/API verification, resource budget review, and rollback guidance.\n3. Run only the smallest reversible verification first: Pull and run one small model on an isolated port, verify health/API behavior, then stop and clean up the runtime..\n4. Keep credentials and primary data out of the first run.\n5. Record output, timeout, cleanup behavior, and missing evidence before adoption.\n\n## Source command candidates\n\n- Python / pip · 官方安装入口: `pip install ollama`\n"
    }
  },
  "validation_id": "dval_79f405cb6efb4ba58f5cb60a06b5c482",
  "locale": "en"
}