可靠推理与自验证：对推理结果做命题抽取、矛盾检测、覆盖度评估与事实锚定校验，输出可验证的结论，纯Python零依赖。当需要"验证推理是否正确""自查答案""检测逻辑矛盾""reason-verify"时使用。
数据分析
llm-verify
试用Verify an LLM API endpoint — model authenticity, billing inflation, relay provenance, performance and silent downgrades. Use when the user asks whether the model they are paying for is genuine, whether a relay or proxy is trustworthy, whether they are being overcharged, or whether a model has been quietly downgraded. 检测 LLM API 端点的真伪、计费掺水、中转来源、性能与降智；当用户问"我用的模型是不是真的"、"这个中转站靠谱吗"、"是不是被降智了"、"计费对不对"时使用。
它能做什么
Verify an LLM API endpoint — model authenticity, billing inflation, relay provenance, performance and silent downgrades. Use when the user asks whether the model they are paying for is genuine, whether a relay or proxy is trustworthy, whether they are being overcharged, or whether a model has been quietly downgraded. 检测 LLM API 端点的真伪、计费掺水、中转来源、性能与降智；当用户问"我用的模型是不是真的"、"这个中转站靠谱吗"、"是不是被降智了"、"计费对不对"时使用。
技能文档
llm-verify
llm-verify is a single-binary CLI that runs a black-box check against any LLM
API endpoint and answers what the user actually wants to know: am I getting the
model I am paying for, am I being overcharged, and how many relays sit on this
path?
When to use this
Reach for this skill when the user asks anything like:
- "Is this relay/proxy trustworthy?" "Is this key really Claude?"
- "Has the model been downgraded?" "It feels dumber than it used to."
- "Is the billing right?" "The token counts look off."
- "Check this API endpoint for me."
- Debugging an endpoint that is slow, drops its stream, or breaks tool calls.
How to run it
llm-verify --base-url --api-key --model
| Flag | Meaning |
|---|---|
--protocol anthropic|openai | Inferred from the URL and model name if omitted |
--depth fast|balanced|forensic | Default balanced; forensic samples more — slower and costlier, but firmer |
--claimed-model | Use when the vendor's advertised name differs from the ID you request |
--lang en|zh | Report language; follows the system locale by default |
-o report.html | HTML report path |
--json report.json | Also emit machine-readable JSON |
--no-open | Do not open a browser |
Credentials can also come from .env or the environment:
LLM_VERIFY_BASE_URL, LLM_VERIFY_API_KEY, LLM_VERIFY_MODEL.
Exit codes
| Code | Meaning |
|---|---|
| 0 | Clean |
| 1 | Failing score, or a suspicious / counterfeit / inconclusive verdict |
| 2 | A hard gate tripped (silent fallback, shared-pool forwarding, tier downgrade, wrapper injection, cache replay, hidden prompt, response replay) |
Suitable as a CI gate as-is.
Reading the result
The tool reports two independent axes. Do not conflate them.
- Authenticity: genuine / genuine-with-defects / relayed / suspicious / counterfeit / inconclusive
- Origin: direct from vendor / cloud platform / subscription-derived / relay / reconstructed channel / undetermined
A real model behind a relay is "relayed", not "counterfeit". That is a longer path, not a substituted model.
Reporting back to the user
- Lead with the conclusion, then the evidence. They want to know whether they can rely on it, not all 40 probe lines.
- Call out hard gates separately. They are facts no weighted score excuses.
- Not tested is not passed. Say plainly which probes were skipped.
- Carry the confidence across, especially for identity: adjacent versions inside one tier are genuinely hard to separate, and the tool abstains when the evidence is thin. Do not supply a verdict it declined to give.
- Do not convict on the tool's behalf. One tier apart is within sampling noise; the tool does not accuse there, and neither should you.
Limits to state honestly
- Resolution stops at tier granularity (flagship / mid / light). Adjacent versions inside a tier cannot be separated without distribution baselines.
- The tier call depends on sampling; use
--depth forensicwhen it matters. - An injected system prompt contaminates identity fingerprints, which is why the contract layer runs first and downgrades identity confidence when it finds injection.
- It cannot prove the server-side weights are the official ones — only that behaviour does or does not match expectations.
- Quantised builds (int4 / fp8) can only be given a probability, never a verdict.
相关技能
AI 产出预检工具。发布博客/提交任务前自动检查：内容完整性、链接可用性、敏感信息泄露、格式合规。受 Claude Code Verify 设计模式启发，完全原创实现，使用 Python + requests。
4 次安装
Check whether this agent is talking to a NEAR AI Cloud TEE endpoint or to an ordinary one, and cryptographically attest that endpoint. Reports which models the chain uses, which ones actually served recent turns, warns when a fallback sends prompts outside the enclave, verifies the Intel TDX quote, and uses the nearai-verify OpenClaw plugin for message-level proof. Trigger when asked "am I private", "is this a TEE", "prove it", "verify your latest response", "which model answered that", or before handling sensitive data.
1 次安装
业务正确性验证层,LLM输出后独立校验业务规则(28平台内容合规+电商价格风控+风控阈值),校验失败拒绝输出并告警。触发:LLM生成内容后/内容发布前/价格设定后/风控检查 不触发:纯文本生成无校验需求
2 次安装
黑盒检测某个 API 端点背后的模型是否"满血/智力正常"（中转站缩水鉴定）。当用户想验证某中转/代理/上游模型是否被换成弱模型、量化版或被剥离 thinking 时使用；也用于对比两个端点的同一模型。触发词如：满血、缩水、智力测试、鉴定模型、benchmark 这个端点。
事实核查助手免费版,面向个人用户提供基础的事实核查能力。通过对接全球专业事实核查机构,对用户提供的声明进行多源交叉验证。Use when 需要AI模型调用、智能对话、Agent编排、LLM应用时使用。不适用于需要100%确定性的关键决策。Use when 需要AI模型调用、智能对话、Agent编排、LLM应用时使用。不适用于需要100%确定性的关键决策.
1 次安装