[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M73DFXGNN2D5H8SY6J18MW60":3,"skill-same:01M73DFXGNN2D5H8SY6J18MW60":67},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":43,"commitSha":44,"skillDirectory":45,"forksCount":46,"repositoryHtmlUrl":47,"repositoryHomepageUrl":48,"ownerHtmlUrl":49,"ownerBlogUrl":50,"ownerDescription":51,"updatedAt":52,"install":53,"resources":66},"01M73DFXGNN2D5H8SY6J18MW60","langfuse",null,"LLM observability with Langfuse — tracing, evals, prompt management, cost tracking","Horonomos-laps\u002FYana-AI","Horonomos-laps","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F355299122?v=4","01M76B3QPXE88ET2SQPDPVATZT"," Horonomos-laps","01M72ZQRJPCYPSSWT4D00VJR2A",13,"Apache-2.0","https:\u002F\u002Fgithub.com\u002FHoronomos-laps\u002FYana-AI\u002Fblob\u002Fmain\u002Fcore\u002Fskills\u002Flangfuse\u002FSKILL.md",[],[],"2026-11-11 15:11:14",false,"---\nname: langfuse\ndescription: LLM observability with Langfuse — tracing, evals, prompt management, cost tracking\ntriggers:\n  - langfuse\n  - llm observability\n  - llm tracing\n  - prompt management langfuse\n  - trace llm calls\n  - langfuse eval\n  - llm cost tracking\n  - langfuse sdk\n  - observe decorator\n  - langfuse dataset\ndo_not_use_for:\n  - generic logging — use structlog\u002Floguru\n  - application monitoring — use OpenTelemetry\n  - model benchmarks — use ragas\u002Fdeepeval\nsee_also:\n  - ragas\n  - deepeval\n  - litellm\n  - pydantic-ai\n---\n\n# Langfuse — LLM Observability\n\n## Core: Trace + Span\n\n```python\nfrom langfuse import Langfuse\nfrom langfuse.decorators import observe, langfuse_context\n\nlangfuse = Langfuse(\n    public_key=\"pk-lf-...\",\n    secret_key=\"sk-lf-...\",\n    host=\"https:\u002F\u002Fcloud.langfuse.com\",  # or self-hosted\n)\n\n@observe()  # auto-creates trace + span\ndef process_document(doc: str) -> str:\n    result = call_llm(doc)\n    langfuse_context.update_current_observation(\n        input=doc,\n        output=result,\n        metadata={\"doc_length\": len(doc)},\n    )\n    return result\n\n@observe(name=\"pipeline\")\ndef run_pipeline(user_input: str) -> dict:\n    langfuse_context.update_current_trace(\n        user_id=\"user-123\",\n        session_id=\"session-456\",\n        tags=[\"production\", \"v2\"],\n    )\n    extracted = extract(user_input)      # auto-nested span\n    summarized = summarize(extracted)    # auto-nested span\n    return {\"result\": summarized}\n```\n\n## Manual Tracing (SDK)\n\n```python\ntrace = langfuse.trace(\n    name=\"rag-query\",\n    user_id=\"user-123\",\n    input={\"query\": user_query},\n)\n\nretrieval = trace.span(\n    name=\"retrieval\",\n    input={\"query\": user_query},\n)\ndocs = vector_store.search(user_query)\nretrieval.end(output={\"doc_count\": len(docs)})\n\ngeneration = trace.generation(\n    name=\"llm-call\",\n    model=\"claude-sonnet-4-6\",\n    input=[{\"role\": \"user\", \"content\": prompt}],\n    model_parameters={\"temperature\": 0.7},\n)\nresponse = call_claude(prompt)\ngeneration.end(\n    output=response,\n    usage={\"input\": 500, \"output\": 200, \"unit\": \"TOKENS\"},\n)\n\ntrace.update(output={\"answer\": response})\nlangfuse.flush()  # ensure all events sent before process exit\n```\n\n## Prompt Management\n\n```python\n# Fetch versioned prompt from Langfuse UI\nprompt_obj = langfuse.get_prompt(\"rag-system-prompt\", version=3)\ncompiled = prompt_obj.compile(\n    context=\"{{context}}\",\n    question=\"{{question}}\",\n)\n\n# Use in generation — links prompt version to trace\ngeneration = trace.generation(\n    name=\"llm-call\",\n    prompt=prompt_obj,  # links version automatically\n    input=compiled,\n)\n```\n\n## Evals (Scores)\n\n```python\n# Manual score after human review\nlangfuse.score(\n    trace_id=trace.id,\n    name=\"faithfulness\",\n    value=0.92,                         # 0.0–1.0\n    comment=\"All claims backed by docs\",\n)\n\n# LLM-as-judge eval\nfrom langfuse.model_based_eval import evaluate_with_llm\n\nscore = evaluate_with_llm(\n    trace_id=trace.id,\n    evaluator=\"hallucination\",          # built-in evaluator\n)\n\n# Python callback for custom eval\ndef score_relevance(trace_id: str, input: str, output: str) -> float:\n    prompt = f\"Rate relevance 0-1: Q={input} A={output}\"\n    return float(call_llm(prompt))\n```\n\n## Datasets & Experiments\n\n```python\n# Create dataset\ndataset = langfuse.create_dataset(name=\"rag-test-set\")\nlangfuse.create_dataset_item(\n    dataset_name=\"rag-test-set\",\n    input={\"query\": \"What is RAG?\"},\n    expected_output={\"answer\": \"Retrieval-Augmented Generation...\"},\n)\n\n# Run experiment over dataset\nitems = langfuse.get_dataset(\"rag-test-set\").items\nfor item in items:\n    with item.observe(run_name=\"v2-experiment\") as trace:\n        output = my_pipeline(item.input[\"query\"])\n        trace.score(name=\"correctness\", value=score(output, item.expected_output))\n```\n\n## LangChain \u002F LlamaIndex Integration\n\n```python\n# LangChain — one-line integration\nfrom langfuse.callback import CallbackHandler\n\nhandler = CallbackHandler(\n    public_key=\"pk-lf-...\",\n    secret_key=\"sk-lf-...\",\n    session_id=\"session-123\",\n)\nchain.invoke({\"input\": query}, config={\"callbacks\": [handler]})\n\n# LlamaIndex\nfrom llama_index.callbacks.langfuse import LangfuseCallbackHandler\nimport llama_index\n\nllama_index.global_handler = LangfuseCallbackHandler()\n```\n\n## Cost Tracking\n\n```python\n# Costs are auto-computed from usage + model pricing\ngeneration.end(\n    output=response_text,\n    usage={\n        \"input\": prompt_tokens,\n        \"output\": completion_tokens,\n        \"unit\": \"TOKENS\",           # or CHARACTERS, MILLISECONDS\n        \"input_cost\": 0.003,        # override if custom model\n        \"output_cost\": 0.015,\n    },\n)\n# View totals in Langfuse dashboard: \u002Fdashboard\u002Fcost\n```\n\n## Anti-Fake-Pass Checks\n\n- `langfuse.flush()` before process exit — otherwise buffered events lost\n- `@observe()` requires `LANGFUSE_PUBLIC_KEY` + `LANGFUSE_SECRET_KEY` env vars or explicit init\n- Trace IDs are UUIDs — store them to link scores back later\n- `prompt_obj.compile()` raises `KeyError` if template variable missing from kwargs\n- `get_dataset().items` is paginated — iterate with `while True` + `next_page` for large sets\n","# Langfuse — LLM Observability\n\n## Core: Trace + Span\n\n```python\nfrom langfuse import Langfuse\nfrom langfuse.decorators import observe, langfuse_context\n\nlangfuse = Langfuse(\n    public_key=\"pk-lf-...\",\n    secret_key=\"sk-lf-...\",\n    host=\"https:\u002F\u002Fcloud.langfuse.com\",  # or self-hosted\n)\n\n@observe()  # auto-creates trace + span\ndef process_document(doc: str) -> str:\n    result = call_llm(doc)\n    langfuse_context.update_current_observation(\n        input=doc,\n        output=result,\n        metadata={\"doc_length\": len(doc)},\n    )\n    return result\n\n@observe(name=\"pipeline\")\ndef run_pipeline(user_input: str) -> dict:\n    langfuse_context.update_current_trace(\n        user_id=\"user-123\",\n        session_id=\"session-456\",\n        tags=[\"production\", \"v2\"],\n    )\n    extracted = extract(user_input)      # auto-nested span\n    summarized = summarize(extracted)    # auto-nested span\n    return {\"result\": summarized}\n```\n\n## Manual Tracing (SDK)\n\n```python\ntrace = langfuse.trace(\n    name=\"rag-query\",\n    user_id=\"user-123\",\n    input={\"query\": user_query},\n)\n\nretrieval = trace.span(\n    name=\"retrieval\",\n    input={\"query\": user_query},\n)\ndocs = vector_store.search(user_query)\nretrieval.end(output={\"doc_count\": len(docs)})\n\ngeneration = trace.generation(\n    name=\"llm-call\",\n    model=\"claude-sonnet-4-6\",\n    input=[{\"role\": \"user\", \"content\": prompt}],\n    model_parameters={\"temperature\": 0.7},\n)\nresponse = call_claude(prompt)\ngeneration.end(\n    output=response,\n    usage={\"input\": 500, \"output\": 200, \"unit\": \"TOKENS\"},\n)\n\ntrace.update(output={\"answer\": response})\nlangfuse.flush()  # ensure all events sent before process exit\n```\n\n## Prompt Management\n\n```python\n# Fetch versioned prompt from Langfuse UI\nprompt_obj = langfuse.get_prompt(\"rag-system-prompt\", version=3)\ncompiled = prompt_obj.compile(\n    context=\"{{context}}\",\n    question=\"{{question}}\",\n)\n\n# Use in generation — links prompt version to trace\ngeneration = trace.generation(\n    name=\"llm-call\",\n    prompt=prompt_obj,  # links version automatically\n    input=compiled,\n)\n```\n\n## Evals (Scores)\n\n```python\n# Manual score after human review\nlangfuse.score(\n    trace_id=trace.id,\n    name=\"faithfulness\",\n    value=0.92,                         # 0.0–1.0\n    comment=\"All claims backed by docs\",\n)\n\n# LLM-as-judge eval\nfrom langfuse.model_based_eval import evaluate_with_llm\n\nscore = evaluate_with_llm(\n    trace_id=trace.id,\n    evaluator=\"hallucination\",          # built-in evaluator\n)\n\n# Python callback for custom eval\ndef score_relevance(trace_id: str, input: str, output: str) -> float:\n    prompt = f\"Rate relevance 0-1: Q={input} A={output}\"\n    return float(call_llm(prompt))\n```\n\n## Datasets & Experiments\n\n```python\n# Create dataset\ndataset = langfuse.create_dataset(name=\"rag-test-set\")\nlangfuse.create_dataset_item(\n    dataset_name=\"rag-test-set\",\n    input={\"query\": \"What is RAG?\"},\n    expected_output={\"answer\": \"Retrieval-Augmented Generation...\"},\n)\n\n# Run experiment over dataset\nitems = langfuse.get_dataset(\"rag-test-set\").items\nfor item in items:\n    with item.observe(run_name=\"v2-experiment\") as trace:\n        output = my_pipeline(item.input[\"query\"])\n        trace.score(name=\"correctness\", value=score(output, item.expected_output))\n```\n\n## LangChain \u002F LlamaIndex Integration\n\n```python\n# LangChain — one-line integration\nfrom langfuse.callback import CallbackHandler\n\nhandler = CallbackHandler(\n    public_key=\"pk-lf-...\",\n    secret_key=\"sk-lf-...\",\n    session_id=\"session-123\",\n)\nchain.invoke({\"input\": query}, config={\"callbacks\": [handler]})\n\n# LlamaIndex\nfrom llama_index.callbacks.langfuse import LangfuseCallbackHandler\nimport llama_index\n\nllama_index.global_handler = LangfuseCallbackHandler()\n```\n\n## Cost Tracking\n\n```python\n# Costs are auto-computed from usage + model pricing\ngeneration.end(\n    output=response_text,\n    usage={\n        \"input\": prompt_tokens,\n        \"output\": completion_tokens,\n        \"unit\": \"TOKENS\",           # or CHARACTERS, MILLISECONDS\n        \"input_cost\": 0.003,        # override if custom model\n        \"output_cost\": 0.015,\n    },\n)\n# View totals in Langfuse dashboard: \u002Fdashboard\u002Fcost\n```\n\n## Anti-Fake-Pass Checks\n\n- `langfuse.flush()` before process exit — otherwise buffered events lost\n- `@observe()` requires `LANGFUSE_PUBLIC_KEY` + `LANGFUSE_SECRET_KEY` env vars or explicit init\n- Trace IDs are UUIDs — store them to link scores back later\n- `prompt_obj.compile()` raises `KeyError` if template variable missing from kwargs\n- `get_dataset().items` is paginated — iterate with `while True` + `next_page` for large sets\n",{"name":5,"see_also":24,"triggers":29,"description":7,"do_not_use_for":39},[25,26,27,28],"ragas","deepeval","litellm","pydantic-ai",[5,30,31,32,33,34,35,36,37,38],"llm observability","llm tracing","prompt management langfuse","trace llm calls","langfuse eval","llm cost tracking","langfuse sdk","observe decorator","langfuse dataset",[40,41,42],"generic logging — use structlog\u002Floguru","application monitoring — use OpenTelemetry","model benchmarks — use ragas\u002Fdeepeval","main","481dcd7dd2e288b9d627a2bd017db70e64e6c6fa","core\u002Fskills\u002Flangfuse",2,"https:\u002F\u002Fgithub.com\u002FHoronomos-laps\u002FYana-AI","https:\u002F\u002Fyana.vutam.link\u002F","https:\u002F\u002Fgithub.com\u002FHoronomos-laps","https:\u002F\u002Fyana.vutam.link","An open-source governed runtime & execution environment for multi-agent AI systems.","2026-11-14 13:24:09",{"source":8,"skillName":5,"commands":54,"prompt":64,"zipAvailable":65},[55,58,61],{"tool":56,"command":57},"npx","npx skills add Horonomos-laps\u002FYana-AI --skill langfuse",{"tool":59,"command":60},"bunx","bunx skills add Horonomos-laps\u002FYana-AI --skill langfuse",{"tool":62,"command":63},"pnpm","pnpm dlx skills add Horonomos-laps\u002FYana-AI --skill langfuse","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FHoronomos-laps\u002FYana-AI\u002Fblob\u002Fmain\u002Fcore\u002Fskills\u002Flangfuse\u002FSKILL.md",true,[],[68,76,83,91,99,107],{"skillNo":69,"name":70,"nameZh":6,"description":71,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":72,"licenseFileUrl":6,"sourceHtmlUrl":73,"categories":74,"keywords":75,"publishedAt":19},"01M735GKET5YVYKAY62W2P5YXB","openai--sentry--sentry","Use when the user asks to inspect Sentry issues or events, summarize recent production errors, or pull basic Sentry health data via the Sentry API; perform read-only queries with the bundled script and require `SENTRY_AUTH_TOKEN`.","MIT","https:\u002F\u002Fgithub.com\u002FHoronomos-laps\u002FYana-AI\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fopenai--sentry--sentry\u002FSKILL.md",[],[],{"skillNo":77,"name":78,"nameZh":6,"description":79,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":80,"categories":81,"keywords":82,"publishedAt":19},"01M73DCAP83T17HY7SMGK7MGFE","ito-basket-compare","Compare Itô prediction-market baskets against a user's knowledge base, portfolio notes, financial context, watchlist, or research thesis. Use for read-only basket comparison and gap analysis without investment advice or live trading.","https:\u002F\u002Fgithub.com\u002FHoronomos-laps\u002FYana-AI\u002Fblob\u002Fmain\u002Fcore\u002Fskills\u002Fito-basket-compare\u002FSKILL.md",[],[],{"skillNo":84,"name":85,"nameZh":6,"description":86,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":87,"sourceHtmlUrl":88,"categories":89,"keywords":90,"publishedAt":19},"01M734CYEMZFDR3KH6SW0JP8TD","implementing-network-traffic-baselining","Build network traffic baselines from NetFlow\u002FIPFIX data using Python pandas for statistical analysis, z-score anomaly detection, and hourly\u002Fdaily traffic pattern profiling","https:\u002F\u002Fgithub.com\u002Fyanacuti1121\u002FYana-AI\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fimplementing-network-traffic-baselining\u002FLICENSE","https:\u002F\u002Fgithub.com\u002FHoronomos-laps\u002FYana-AI\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fimplementing-network-traffic-baselining\u002FSKILL.md",[],[],{"skillNo":92,"name":93,"nameZh":6,"description":94,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":95,"sourceHtmlUrl":96,"categories":97,"keywords":98,"publishedAt":19},"01M73ASZ6NJRJM44Z2P84NGQ3Q","detecting-evasion-techniques-in-endpoint-logs","Detects defense evasion techniques used by adversaries in endpoint logs including log tampering, timestomping, process injection, and security tool disabling. Use when investigating suspicious endpoint behavior, building detection rules for evasion tactics, or conducting threat hunting for stealthy adversary activity. Activates for requests involving evasion detection, defense evasion analysis, log tampering detection, or MITRE ATT&CK TA0005.","https:\u002F\u002Fgithub.com\u002Fyanacuti1121\u002FYana-AI\u002Fblob\u002Fmain\u002Fcore\u002Fskills\u002Fdetecting-evasion-techniques-in-endpoint-logs\u002FLICENSE","https:\u002F\u002Fgithub.com\u002FHoronomos-laps\u002FYana-AI\u002Fblob\u002Fmain\u002Fcore\u002Fskills\u002Fdetecting-evasion-techniques-in-endpoint-logs\u002FSKILL.md",[],[],{"skillNo":100,"name":101,"nameZh":6,"description":102,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":103,"sourceHtmlUrl":104,"categories":105,"keywords":106,"publishedAt":19},"01M734AFNXD06Z78QRK0VKGT7Q","implementing-microsegmentation-with-guardicore","Implementing microsegmentation using Akamai Guardicore Segmentation to map application dependencies, create granular network policies, visualize east-west traffic flows, and enforce least-privilege communication between workloads across data centers and cloud.","https:\u002F\u002Fgithub.com\u002Fyanacuti1121\u002FYana-AI\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fimplementing-microsegmentation-with-guardicore\u002FLICENSE","https:\u002F\u002Fgithub.com\u002FHoronomos-laps\u002FYana-AI\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fimplementing-microsegmentation-with-guardicore\u002FSKILL.md",[],[],{"skillNo":108,"name":109,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":110,"categories":111,"keywords":112,"publishedAt":19},"01M736TK27QMCXK0DVCH25SJN5","ramdisk-tmpfs-patterns","https:\u002F\u002Fgithub.com\u002FHoronomos-laps\u002FYana-AI\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Framdisk-tmpfs-patterns\u002FSKILL.md",[],[]]