[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M73DXK0C9ZQS6SWVFT83H0DG":3,"skill-same:01M73DXK0C9ZQS6SWVFT83H0DG":47},{"skillNo":4,"name":5,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"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":30,"ownerDescription":31,"updatedAt":32,"install":33,"resources":46},"01M73DXK0C9ZQS6SWVFT83H0DG","needle-haystack-context",null,"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\u002Fneedle-haystack-context\u002FSKILL.md",[],[],"2026-11-11 15:11:14",false,"---\nname: needle-haystack-context\ndescription: Needle-in-a-Haystack context placement testing for LLM retrieval accuracy. Context window stress tests, information placement strategies, lost-in-the-middle avoidance, and optimal document ordering for agent prompts. Sources: gkamradt\u002FLLMTest_NeedleInAHaystack (MIT).\norigin: yana-ai — synthesized from gkamradt\u002FLLMTest_NeedleInAHaystack (MIT)\nlicense: Apache-2.0\nversion: 1.0.0\ncompatibility: yana-ai >= 1.3.51\n---\n\n# \u002Fneedle-haystack-context\n\n## When to Use\n\n- Test whether an LLM can retrieve a specific fact buried in a long context window\n- Choose optimal placement for critical information in agent system prompts\n- Diagnose why an agent \"forgets\" instructions in long conversations\n- Benchmark model context quality before deploying for long-context RAG tasks\n\n## Do NOT use for\n\n- Short contexts \u003C 4k tokens (all models handle these reliably)\n- Semantic search (use [[in-memory-vector-storage]] for retrieval)\n\n---\n\n## Lost-in-the-Middle effect\n\n```\nResearch finding (Liu et al. 2023):\n  Models perform best when critical info is at START or END of context.\n  Performance degrades ~40% when critical info is buried in the MIDDLE.\n\n  0%──────────────────────100% position in context\n  ████████░░░░░░░░░░░░████████  ← retrieval accuracy\n  Best    Worst     Best\n\nAgent implication: put rules\u002Fconstraints at the TOP of system prompt,\n  not buried in the middle of a long document list.\n```\n\n---\n\n## Needle-in-Haystack test runner\n\n```python\nimport anthropic\n\ndef needle_test(\n  client:    anthropic.Anthropic,\n  needle:    str,   # fact to retrieve: \"The magic number is 42\"\n  haystack:  str,   # long filler text (Paul Graham essays, etc.)\n  position:  float, # 0.0 = start, 0.5 = middle, 1.0 = end\n  ctx_len:   int,   # target context length in tokens\n) -> bool:\n  # Place needle at the specified position\n  words   = haystack.split()\n  insert  = int(len(words) * position)\n  padded  = ' '.join(words[:insert]) + f'\\n\\n{needle}\\n\\n' + ' '.join(words[insert:])\n  trimmed = padded[:ctx_len * 4]  # ~4 chars\u002Ftoken\n\n  response = client.messages.create(\n    model      = 'claude-sonnet-4-6',\n    max_tokens = 64,\n    system     = 'Answer based only on the provided text.',\n    messages   = [\n      { 'role': 'user', 'content': f'{trimmed}\\n\\nWhat is the magic number?' }\n    ],\n  )\n  answer = response.content[0].text\n  return '42' in answer\n```\n\n---\n\n## Sweep test across positions and context lengths\n\n```python\nimport numpy as np\n\npositions = [0.0, 0.1, 0.25, 0.5, 0.75, 0.9, 1.0]\nctx_lens  = [2048, 4096, 8192, 16384]\nresults   = {}\n\nfor ctx in ctx_lens:\n  for pos in positions:\n    hit = needle_test(client, needle, haystack, pos, ctx)\n    results[(ctx, pos)] = hit\n    print(f'ctx={ctx:6d} pos={pos:.1f} → {\"PASS\" if hit else \"FAIL\"}')\n\n# Identify weak spots\nfailures = [(ctx, pos) for (ctx, pos), ok in results.items() if not ok]\nprint('\\nFailed positions:', failures)\n```\n\n---\n\n## Agent prompt ordering rules (from NIAH findings)\n\n```\nPriority order for system prompt sections:\n  1. [TOP]    Hard constraints, safety rules, tool restrictions\n  2. [TOP]    Agent identity and role\n  3. [MIDDLE] Reference documents, context (expendable)\n  4. [BOTTOM] Specific task instructions\n  5. [BOTTOM] Output format requirements\n\nNEVER bury \"do not do X\" rules in the middle of a long document list.\nIf context > 32k tokens: summarize middle sections, keep rules at edges.\n```\n\n---\n\n## Anti-Fake-Pass Checklist\n\n```\n❌ Testing with model's advertised context length but not checking quality → model \"accepts\" 128k but retrieves poorly past 32k\n❌ Needle is too similar to haystack content → false positives (model guesses correctly without retrieval)\n❌ Single position test → must sweep all positions to find the degradation curve\n❌ Counting tokens by character\u002F4 for needle placement → use tiktoken for accurate position\n❌ Using this test for semantic understanding → NIAH tests literal retrieval only\n❌ Fixing NIAH failures by just shortening context → also restructure prompt (put key info at top)\n```\n","# \u002Fneedle-haystack-context\n\n## When to Use\n\n- Test whether an LLM can retrieve a specific fact buried in a long context window\n- Choose optimal placement for critical information in agent system prompts\n- Diagnose why an agent \"forgets\" instructions in long conversations\n- Benchmark model context quality before deploying for long-context RAG tasks\n\n## Do NOT use for\n\n- Short contexts \u003C 4k tokens (all models handle these reliably)\n- Semantic search (use [[in-memory-vector-storage]] for retrieval)\n\n---\n\n## Lost-in-the-Middle effect\n\n```\nResearch finding (Liu et al. 2023):\n  Models perform best when critical info is at START or END of context.\n  Performance degrades ~40% when critical info is buried in the MIDDLE.\n\n  0%──────────────────────100% position in context\n  ████████░░░░░░░░░░░░████████  ← retrieval accuracy\n  Best    Worst     Best\n\nAgent implication: put rules\u002Fconstraints at the TOP of system prompt,\n  not buried in the middle of a long document list.\n```\n\n---\n\n## Needle-in-Haystack test runner\n\n```python\nimport anthropic\n\ndef needle_test(\n  client:    anthropic.Anthropic,\n  needle:    str,   # fact to retrieve: \"The magic number is 42\"\n  haystack:  str,   # long filler text (Paul Graham essays, etc.)\n  position:  float, # 0.0 = start, 0.5 = middle, 1.0 = end\n  ctx_len:   int,   # target context length in tokens\n) -> bool:\n  # Place needle at the specified position\n  words   = haystack.split()\n  insert  = int(len(words) * position)\n  padded  = ' '.join(words[:insert]) + f'\\n\\n{needle}\\n\\n' + ' '.join(words[insert:])\n  trimmed = padded[:ctx_len * 4]  # ~4 chars\u002Ftoken\n\n  response = client.messages.create(\n    model      = 'claude-sonnet-4-6',\n    max_tokens = 64,\n    system     = 'Answer based only on the provided text.',\n    messages   = [\n      { 'role': 'user', 'content': f'{trimmed}\\n\\nWhat is the magic number?' }\n    ],\n  )\n  answer = response.content[0].text\n  return '42' in answer\n```\n\n---\n\n## Sweep test across positions and context lengths\n\n```python\nimport numpy as np\n\npositions = [0.0, 0.1, 0.25, 0.5, 0.75, 0.9, 1.0]\nctx_lens  = [2048, 4096, 8192, 16384]\nresults   = {}\n\nfor ctx in ctx_lens:\n  for pos in positions:\n    hit = needle_test(client, needle, haystack, pos, ctx)\n    results[(ctx, pos)] = hit\n    print(f'ctx={ctx:6d} pos={pos:.1f} → {\"PASS\" if hit else \"FAIL\"}')\n\n# Identify weak spots\nfailures = [(ctx, pos) for (ctx, pos), ok in results.items() if not ok]\nprint('\\nFailed positions:', failures)\n```\n\n---\n\n## Agent prompt ordering rules (from NIAH findings)\n\n```\nPriority order for system prompt sections:\n  1. [TOP]    Hard constraints, safety rules, tool restrictions\n  2. [TOP]    Agent identity and role\n  3. [MIDDLE] Reference documents, context (expendable)\n  4. [BOTTOM] Specific task instructions\n  5. [BOTTOM] Output format requirements\n\nNEVER bury \"do not do X\" rules in the middle of a long document list.\nIf context > 32k tokens: summarize middle sections, keep rules at edges.\n```\n\n---\n\n## Anti-Fake-Pass Checklist\n\n```\n❌ Testing with model's advertised context length but not checking quality → model \"accepts\" 128k but retrieves poorly past 32k\n❌ Needle is too similar to haystack content → false positives (model guesses correctly without retrieval)\n❌ Single position test → must sweep all positions to find the degradation curve\n❌ Counting tokens by character\u002F4 for needle placement → use tiktoken for accurate position\n❌ Using this test for semantic understanding → NIAH tests literal retrieval only\n❌ Fixing NIAH failures by just shortening context → also restructure prompt (put key info at top)\n```\n",{},"main","481dcd7dd2e288b9d627a2bd017db70e64e6c6fa","core\u002Fskills\u002Fneedle-haystack-context",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":7,"skillName":5,"commands":34,"prompt":44,"zipAvailable":45},[35,38,41],{"tool":36,"command":37},"npx","npx skills add Horonomos-laps\u002FYana-AI --skill needle-haystack-context",{"tool":39,"command":40},"bunx","bunx skills add Horonomos-laps\u002FYana-AI --skill needle-haystack-context",{"tool":42,"command":43},"pnpm","pnpm dlx skills add Horonomos-laps\u002FYana-AI --skill needle-haystack-context","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FHoronomos-laps\u002FYana-AI\u002Fblob\u002Fmain\u002Fcore\u002Fskills\u002Fneedle-haystack-context\u002FSKILL.md",true,[],[48,56,63,71,79,87],{"skillNo":49,"name":50,"nameZh":6,"description":51,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":52,"licenseFileUrl":6,"sourceHtmlUrl":53,"categories":54,"keywords":55,"publishedAt":18},"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":57,"name":58,"nameZh":6,"description":59,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":60,"categories":61,"keywords":62,"publishedAt":18},"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":64,"name":65,"nameZh":6,"description":66,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":67,"sourceHtmlUrl":68,"categories":69,"keywords":70,"publishedAt":18},"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":72,"name":73,"nameZh":6,"description":74,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":75,"sourceHtmlUrl":76,"categories":77,"keywords":78,"publishedAt":18},"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":80,"name":81,"nameZh":6,"description":82,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":83,"sourceHtmlUrl":84,"categories":85,"keywords":86,"publishedAt":18},"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":88,"name":89,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":90,"categories":91,"keywords":92,"publishedAt":18},"01M736TK27QMCXK0DVCH25SJN5","ramdisk-tmpfs-patterns","https:\u002F\u002Fgithub.com\u002FHoronomos-laps\u002FYana-AI\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Framdisk-tmpfs-patterns\u002FSKILL.md",[],[]]