)
### Product Analytics · openai/gpt-4.1-mini
keywordId: watch-keyword-532e310d54beebe2356227d4 · runId: 942c0129-c245-4a00-9810-3f8bcde4d518 · probeId: d85e30b0-0dd7-46d9-900c-7c0cb22273e4
provider_native · failed · firstAttemptId: 829eb19d-0fb0-4d0d-9055-1a7e62188ffd
analysisStatus: analysis_failed · resultAttemptId: 829eb19d-0fb0-4d0d-9055-1a7e62188ffd
- mentionJudgment: unknown
- recommendationJudgment: unknown
无法确认: Unterminated string in JSON at position 4368 (line 1 column 4369)
[实际请求与原文证据](./public-evidence.json)
#### Attempt 829eb19d-0fb0-4d0d-9055-1a7e62188ffd
analysis_failed · 时间: 2026-09-08T06:05:50.138Z · 首次 Attempt
requestedSearch: provider_native · used: true · executionMode: native
OpenRouter metadata confirmed provider-native web search.
错误: analysis_failed · Unterminated string in JSON at position 4368 (line 1 column 4369)
finish_reason: stop
查看模型原始回答
{"analysisStatus":"completed","mentions":[{"name":"Pendo","domain":"pendo.io","recommendation":"positive","mentionQuote":"Pendo's analytics tool that instantly tracks what drives user engagement, retention, and product success—no manual event tagging needed.","recommendationQuote":"Pendo has been a huge win for us in this time when every dollar counts and efficiency is super important. It’s made collaboration more grounded in tangible objectives and results, and it provides metrics that define what ‘better’ really means.","firstMentionOffset":0,"firstRecommendationOffset":0,"firstMentionState":"unique","firstRecommendationState":"unique"},{"name":"Amplitude","domain":"amplitude.com","recommendation":"positive","mentionQuote":"Amplitude is a powerful product intelligence platform that will empower your product managers to self-serve and, in turn, allow you to make better and faster product decisions.","recommendationQuote":"Product intelligence is not simply reporting numbers. It’s about helping the business make measurably better decisions.","firstMentionOffset":0,"firstRecommendationOffset":0,"firstMentionState":"unique","firstRecommendationState":"unique"},{"name":"Whatfix","domain":"whatfix.com","recommendation":"positive","mentionQuote":"Whatfix Product Analytics goes beyond traditional dashboards by analyzing every user journey to reveal the ‘why’ behind the data. Pinpoint friction, fix issues faster, and deliver targeted in-app experiences that boost adoption and retention.","recommendationQuote":"Our Insights Agent analyzes user behavior and application events in real time, delivering clear explanations and tailored recommendations so your team can move faster, smarter, and with confidence.","firstMentionOffset":0,"firstRecommendationOffset":0,"firstMentionState":"unique","firstRecommendationState":"unique"},{"name":"Google Analytics","domain":"analytics.google.com","recommendation":"positive","mentionQuote":"Google Analytics is our overall pick for product analytics in 2026.","recommendationQuote":"Google Analytics is our #1 pick for product analytics for enterprises in 2026.","firstMentionOffset":0,"firstRecommendationOffset":0,"firstMentionState":"unique","firstRecommendationState":"unique"},{"name":"LogRocket","domain":"logrocket.com","recommendation":"positive","mentionQuote":"LogRocket is our overall pick for product analytics in 2026.","recommendationQuote":"LogRocket is our #1 pick for product analytics for enterprises in 2026.","firstMentionOffset":0,"firstRecommendationOffset":0,"firstMentionState":"unique","firstRecommendationState":"unique"},{"name":"Mixpanel","domain":"mixpanel.com","recommendation":"positive","mentionQuote":"Mixpanel is our overall pick for product analytics in 2026.","recommendationQuote":"Mixpanel is our #1 pick for product analytics for enterprises in 2026.","firstMentionOffset":0,"firstRecommendationOffset":0,"firstMentionState":"unique","firstRecommendationState":"unique"},{"name":"PostHog","domain":"posthog.com","recommendation":"positive","mentionQuote":"PostHog is our overall pick for product analytics in 2026.","recommendationQuote":"PostHog is our #1 pick for product analytics for enterprises in 2026.","firstMentionOffset":0,"firstRecommendationOffset":0,"firstMentionState":"unique","firstRecommendationState":"unique"},{"name":"Heap","domain":"heap.io","recommendation":"positive","mentionQuote":"Heap is our overall pick for product analytics in 2026.","recommendationQuote":"Heap is our #1 pick for product analytics for enterprises in 2026.","firstMentionOffset":0,"firstRecommendationOffset":0,"firstMentionState":"unique","firstRecommendationState":"unique"},{"name":"ChurnZero","domain":"churnzero.com","recommendation":"positive","mentionQuote":"ChurnZero is our overall pick for product analytics in 2026.","recommendationQuote":"ChurnZero is our #1 pick for product analytics for enterprises in 2026.","firstMentionOffset":0,"firstRecommendationOffset":0,"firstMentionState":"unique","firstRecommendationState":"unique"},{"name":"mParticle","domain":"mparticle.com","recommendation":"positive","mentionQuote":"mParticle is our overall pick for product analytics in 2026.","recommendationQuote":"mParticle is our #1 pick for product analytics for enterprises in 2026.","firstMentionOffset":0,"firstRecommendationOffset":0,"firstMentionState":"unique","
SHA-256: `307ab966b57764ddc7f8a252ba9a8009d2d407f736761cb36dd16736f5bf9d98`
#### Provider 引用
此 Attempt 未归档此类来源。
#### 搜索检索结果
此 Attempt 未归档此类来源。
#### 回答中的链接(非 Provider 引用)
此 Attempt 未归档正文 URL。
### Feature Flags · openai/gpt-4o-mini
keywordId: watch-keyword-058bb884e2e3e12f02113bfd · runId: 942c0129-c245-4a00-9810-3f8bcde4d518 · probeId: 6d6ceff5-1d01-4ec4-9632-ec3f825def67
off · completed · firstAttemptId: 4e691257-96e7-4dc2-867f-39a149f0ab76
analysisStatus: completed · resultAttemptId: 4e691257-96e7-4dc2-867f-39a149f0ab76
- mentionJudgment: adjudicable
- recommendationJudgment: adjudicable
- Feature Flags: mentioned · mention: Feature Flags are a powerful tool for managing software features and deployments. · recommendation: — · attemptId: 4e691257-96e7-4dc2-867f-39a149f0ab76
无法确认: —
[实际请求与原文证据](./public-evidence.json)
#### Attempt 4e691257-96e7-4dc2-867f-39a149f0ab76
completed · 时间: 2026-09-08T06:05:35.479Z · 首次 Attempt
requestedSearch: off · used: false · executionMode: unverified
finish_reason: stop
查看模型原始回答
{"analysisStatus":"completed","mentions":[{"name":"Feature Flags","domain":null,"recommendation":"mentioned","mentionQuote":"Feature Flags are a powerful tool for managing software features and deployments.","recommendationQuote":null,"firstMentionOffset":0,"firstRecommendationOffset":null,"firstMentionState":"unique","firstRecommendationState":"none"}],"unknowns":[]}
SHA-256: `50f2c358be54909aed7f990b6d7d8b096d50c5c27806912a4769d684eb826cb2`
#### Provider 引用
此 Attempt 未归档此类来源。
#### 搜索检索结果
此 Attempt 未归档此类来源。
#### 回答中的链接(非 Provider 引用)
此 Attempt 未归档正文 URL。
### Product Analytics · openai/gpt-4o-mini
keywordId: watch-keyword-532e310d54beebe2356227d4 · runId: 942c0129-c245-4a00-9810-3f8bcde4d518 · probeId: 5ab15ac9-aa9d-48ed-9e93-8cc13958e0f2
off · completed · firstAttemptId: 6b06afab-4ad5-454b-9c58-6b0119d45276
analysisStatus: completed · resultAttemptId: 6b06afab-4ad5-454b-9c58-6b0119d45276
- mentionJudgment: adjudicable
- recommendationJudgment: adjudicable
- Product Analytics: mentioned · mention: Product Analytics is a crucial aspect of understanding user behavior and improving product performance. · recommendation: — · attemptId: 6b06afab-4ad5-454b-9c58-6b0119d45276
无法确认: —
[实际请求与原文证据](./public-evidence.json)
#### Attempt 6b06afab-4ad5-454b-9c58-6b0119d45276
completed · 时间: 2026-09-08T06:05:47.503Z · 首次 Attempt
requestedSearch: off · used: false · executionMode: unverified
finish_reason: stop
查看模型原始回答
{"analysisStatus":"completed","mentions":[{"name":"Product Analytics","domain":null,"recommendation":"mentioned","mentionQuote":"Product Analytics is a crucial aspect of understanding user behavior and improving product performance.","recommendationQuote":null,"firstMentionOffset":0,"firstRecommendationOffset":null,"firstMentionState":"unique","firstRecommendationState":"none"}],"unknowns":[]}
SHA-256: `640b72a178649ee1457cab4da0a4176b38539500ae8d976f18de144f5e8da382`
#### Provider 引用
此 Attempt 未归档此类来源。
#### 搜索检索结果
此 Attempt 未归档此类来源。
#### 回答中的链接(非 Provider 引用)
此 Attempt 未归档正文 URL。
### Feature Flags · google/gemini-2.5-flash-lite
keywordId: watch-keyword-058bb884e2e3e12f02113bfd · runId: 942c0129-c245-4a00-9810-3f8bcde4d518 · probeId: 4ebc113e-b3f7-4a98-96b2-0d9e0702dae0
off · completed · firstAttemptId: c8c49cee-00b8-4fa3-9df8-0f9a0dcdcd8e
analysisStatus: completed · resultAttemptId: c8c49cee-00b8-4fa3-9df8-0f9a0dcdcd8e
- mentionJudgment: adjudicable
- recommendationJudgment: adjudicable
- Feature Flags: mentioned · mention: Feature Flags · recommendation: — · attemptId: c8c49cee-00b8-4fa3-9df8-0f9a0dcdcd8e
无法确认: —
[实际请求与原文证据](./public-evidence.json)
#### Attempt c8c49cee-00b8-4fa3-9df8-0f9a0dcdcd8e
completed · 时间: 2026-09-08T06:05:46.397Z · 首次 Attempt
requestedSearch: off · used: false · executionMode: unverified
finish_reason: stop
查看模型原始回答
{
"analysisStatus": "completed",
"mentions": [
{
"name": "Feature Flags",
"domain": null,
"recommendation": "mentioned",
"mentionQuote": "Feature Flags",
"recommendationQuote": null,
"firstMentionOffset": 0,
"firstRecommendationOffset": null,
"firstMentionState": "unique",
"firstRecommendationState": "none"
}
],
"unknowns": []
}
SHA-256: `6d346060e85b092bba46e1684aa09c57db313bfc948da3d3e0e27ba0da25cf2e`
#### Provider 引用
此 Attempt 未归档此类来源。
#### 搜索检索结果
此 Attempt 未归档此类来源。
#### 回答中的链接(非 Provider 引用)
此 Attempt 未归档正文 URL。
### Product Analytics · google/gemini-2.5-flash-lite
keywordId: watch-keyword-532e310d54beebe2356227d4 · runId: 942c0129-c245-4a00-9810-3f8bcde4d518 · probeId: 1c0fa6b5-f29f-44d1-8898-050b3a7c7cb6
off · completed · firstAttemptId: 966350ea-f392-4fb0-8998-d108009cbe1b
analysisStatus: completed · resultAttemptId: 966350ea-f392-4fb0-8998-d108009cbe1b
- mentionJudgment: adjudicable
- recommendationJudgment: adjudicable
- Product Analytics: mentioned · mention: Product Analytics · recommendation: — · attemptId: 966350ea-f392-4fb0-8998-d108009cbe1b
无法确认: —
[实际请求与原文证据](./public-evidence.json)
#### Attempt 966350ea-f392-4fb0-8998-d108009cbe1b
completed · 时间: 2026-09-08T06:05:51.531Z · 首次 Attempt
requestedSearch: off · used: false · executionMode: unverified
finish_reason: stop
查看模型原始回答
{
"analysisStatus": "completed",
"mentions": [
{
"name": "Product Analytics",
"domain": null,
"recommendation": "mentioned",
"mentionQuote": "Product Analytics",
"recommendationQuote": null,
"firstMentionOffset": 0,
"firstRecommendationOffset": null,
"firstMentionState": "unique",
"firstRecommendationState": "none"
}
],
"unknowns": []
}
SHA-256: `848c0ad989bb26d0ecb5bfc9f319d02a19ae0e817ad211a7278db240872fd70b`
#### Provider 引用
此 Attempt 未归档此类来源。
#### 搜索检索结果
此 Attempt 未归档此类来源。
#### 回答中的链接(非 Provider 引用)
此 Attempt 未归档正文 URL。
### Feature Flags · openai/gpt-4o-mini
keywordId: watch-keyword-058bb884e2e3e12f02113bfd · runId: 71466c93-74d0-4807-9a42-64bdb7cabc78 · probeId: a85cda5a-2526-4938-b13b-ad1aab33c1e8
off · completed · firstAttemptId: 623918b4-6fdb-4fa9-bf5c-1edf43465d9d
analysisStatus: completed · resultAttemptId: 623918b4-6fdb-4fa9-bf5c-1edf43465d9d
- mentionJudgment: adjudicable
- recommendationJudgment: adjudicable
- Feature Flags: mentioned · mention: Feature Flags are a powerful tool for managing software releases and enabling continuous delivery. · recommendation: — · attemptId: 623918b4-6fdb-4fa9-bf5c-1edf43465d9d
无法确认: —
[实际请求与原文证据](./public-evidence.json)
#### Attempt 623918b4-6fdb-4fa9-bf5c-1edf43465d9d
completed · 时间: 2026-09-08T06:06:03.272Z · 首次 Attempt
requestedSearch: off · used: false · executionMode: unverified
finish_reason: stop
查看模型原始回答
{"analysisStatus":"completed","mentions":[{"name":"Feature Flags","domain":null,"recommendation":"mentioned","mentionQuote":"Feature Flags are a powerful tool for managing software releases and enabling continuous delivery.","recommendationQuote":null,"firstMentionOffset":0,"firstRecommendationOffset":null,"firstMentionState":"unique","firstRecommendationState":"none"}],"unknowns":[]}
SHA-256: `f2fe6c878b421a4105640e5d988e502294061dd241854e42ce6c1221bcde8a4b`
#### Provider 引用
此 Attempt 未归档此类来源。
#### 搜索检索结果
此 Attempt 未归档此类来源。
#### 回答中的链接(非 Provider 引用)
此 Attempt 未归档正文 URL。
### Product Analytics · openai/gpt-4o-mini
keywordId: watch-keyword-532e310d54beebe2356227d4 · runId: 71466c93-74d0-4807-9a42-64bdb7cabc78 · probeId: c7b9f38d-0d97-4a22-86b7-d7ddeba000a8
off · completed · firstAttemptId: dac3b40f-b75a-44b8-a2db-1f9d41fae7d5
analysisStatus: completed · resultAttemptId: dac3b40f-b75a-44b8-a2db-1f9d41fae7d5
- mentionJudgment: adjudicable
- recommendationJudgment: adjudicable
- Product Analytics: mentioned · mention: Product Analytics is a crucial aspect of understanding user behavior and improving product performance. · recommendation: — · attemptId: dac3b40f-b75a-44b8-a2db-1f9d41fae7d5
无法确认: —
[实际请求与原文证据](./public-evidence.json)
#### Attempt dac3b40f-b75a-44b8-a2db-1f9d41fae7d5
completed · 时间: 2026-09-08T06:06:11.725Z · 首次 Attempt
requestedSearch: off · used: false · executionMode: unverified
finish_reason: stop
查看模型原始回答
{"analysisStatus":"completed","mentions":[{"name":"Product Analytics","domain":null,"recommendation":"mentioned","mentionQuote":"Product Analytics is a crucial aspect of understanding user behavior and improving product performance.","recommendationQuote":null,"firstMentionOffset":0,"firstRecommendationOffset":null,"firstMentionState":"unique","firstRecommendationState":"none"}],"unknowns":[]}
SHA-256: `640b72a178649ee1457cab4da0a4176b38539500ae8d976f18de144f5e8da382`
#### Provider 引用
此 Attempt 未归档此类来源。
#### 搜索检索结果
此 Attempt 未归档此类来源。
#### 回答中的链接(非 Provider 引用)
此 Attempt 未归档正文 URL。
### Feature Flags · google/gemini-2.5-flash-lite
keywordId: watch-keyword-058bb884e2e3e12f02113bfd · runId: 71466c93-74d0-4807-9a42-64bdb7cabc78 · probeId: c1edcab9-f83b-42c8-b74b-e753a325284c
off · completed · firstAttemptId: b817cb8a-d2c6-4b16-aa54-dd2d6a8e2156
analysisStatus: completed · resultAttemptId: b817cb8a-d2c6-4b16-aa54-dd2d6a8e2156
- mentionJudgment: adjudicable
- recommendationJudgment: adjudicable
- Feature Flags: mentioned · mention: Feature Flags · recommendation: — · attemptId: b817cb8a-d2c6-4b16-aa54-dd2d6a8e2156
无法确认: —
[实际请求与原文证据](./public-evidence.json)
#### Attempt b817cb8a-d2c6-4b16-aa54-dd2d6a8e2156
completed · 时间: 2026-09-08T06:06:10.644Z · 首次 Attempt
requestedSearch: off · used: false · executionMode: unverified
finish_reason: stop
查看模型原始回答
{
"analysisStatus": "completed",
"mentions": [
{
"name": "Feature Flags",
"domain": null,
"recommendation": "mentioned",
"mentionQuote": "Feature Flags",
"recommendationQuote": null,
"firstMentionOffset": 0,
"firstRecommendationOffset": null,
"firstMentionState": "unique",
"firstRecommendationState": "none"
}
],
"unknowns": []
}
SHA-256: `6d346060e85b092bba46e1684aa09c57db313bfc948da3d3e0e27ba0da25cf2e`
#### Provider 引用
此 Attempt 未归档此类来源。
#### 搜索检索结果
此 Attempt 未归档此类来源。
#### 回答中的链接(非 Provider 引用)
此 Attempt 未归档正文 URL。
### Product Analytics · google/gemini-2.5-flash-lite
keywordId: watch-keyword-532e310d54beebe2356227d4 · runId: 71466c93-74d0-4807-9a42-64bdb7cabc78 · probeId: 996c82f3-42df-4322-bd6a-252bc2f90ba4
off · completed · firstAttemptId: 71774913-d59f-47dc-b509-794f6a4f802a
analysisStatus: completed · resultAttemptId: 71774913-d59f-47dc-b509-794f6a4f802a
- mentionJudgment: adjudicable
- recommendationJudgment: adjudicable
- Product Analytics: mentioned · mention: Product Analytics · recommendation: — · attemptId: 71774913-d59f-47dc-b509-794f6a4f802a
无法确认: —
[实际请求与原文证据](./public-evidence.json)
#### Attempt 71774913-d59f-47dc-b509-794f6a4f802a
completed · 时间: 2026-09-08T06:06:16.275Z · 首次 Attempt
requestedSearch: off · used: false · executionMode: unverified
finish_reason: stop
查看模型原始回答
{
"analysisStatus": "completed",
"mentions": [
{
"name": "Product Analytics",
"domain": null,
"recommendation": "mentioned",
"mentionQuote": "Product Analytics",
"recommendationQuote": null,
"firstMentionOffset": 0,
"firstRecommendationOffset": null,
"firstMentionState": "unique",
"firstRecommendationState": "none"
}
],
"unknowns": []
}
SHA-256: `848c0ad989bb26d0ecb5bfc9f319d02a19ae0e817ad211a7278db240872fd70b`
#### Provider 引用
此 Attempt 未归档此类来源。
#### 搜索检索结果
此 Attempt 未归档此类来源。
#### 回答中的链接(非 Provider 引用)
此 Attempt 未归档正文 URL。
### Feature Flags · openai/gpt-4.1-mini
keywordId: watch-keyword-058bb884e2e3e12f02113bfd · runId: 71466c93-74d0-4807-9a42-64bdb7cabc78 · probeId: dfbba462-d4b5-4144-8d69-96f8bc01d08f
provider_native · completed · firstAttemptId: ea947928-ca00-4b1b-b044-8f4fc6c87e78
analysisStatus: completed · resultAttemptId: ea947928-ca00-4b1b-b044-8f4fc6c87e78
- mentionJudgment: adjudicable
- recommendationJudgment: adjudicable
- Feature Flags: mentioned · mention: Feature flags are a powerful way to fine-tune your control over which features are enabled within a software deployment. ([splunk.com](https://www.splunk.com/en_us/blog/learn/feature-flags.html?utm_source=openai)) · recommendation: — · attemptId: ea947928-ca00-4b1b-b044-8f4fc6c87e78
无法确认: —
[实际请求与原文证据](./public-evidence.json)
#### Attempt ea947928-ca00-4b1b-b044-8f4fc6c87e78
completed · 时间: 2026-09-08T06:06:07.723Z · 首次 Attempt
requestedSearch: provider_native · used: true · executionMode: native
OpenRouter metadata confirmed provider-native web search.
finish_reason: stop
查看模型原始回答
{"analysisStatus":"completed","mentions":[{"name":"Feature Flags","domain":null,"recommendation":"mentioned","mentionQuote":"Feature flags are a powerful way to fine-tune your control over which features are enabled within a software deployment. ([splunk.com](https://www.splunk.com/en_us/blog/learn/feature-flags.html?utm_source=openai))","recommendationQuote":null,"firstMentionOffset":0,"firstRecommendationOffset":0,"firstMentionState":"unique","firstRecommendationState":"none"}],"unknowns":[]}
SHA-256: `a51cb73372d6bd05df42cc0e8a82352786b4447a619eb17cde246a7d058e5b56`
#### Provider 引用
- When and Why To Adopt Feature Flags | Splunk: [https://www.splunk.com/en_us/blog/learn/feature-flags.html?utm_source=openai]() · `choices[0].message.annotations[0].url_citation.url`
#### 搜索检索结果
此 Attempt 未归档此类来源。
#### 回答中的链接(非 Provider 引用)
- [https://www.splunk.com/en_us/blog/learn/feature-flags.html?utm_source=openai]()
### Product Analytics · openai/gpt-4.1-mini
keywordId: watch-keyword-532e310d54beebe2356227d4 · runId: 71466c93-74d0-4807-9a42-64bdb7cabc78 · probeId: 2713f3dd-9000-4378-8f92-36c029d3dae3
provider_native · failed · firstAttemptId: 8371777c-b23e-4b35-b46d-1e02a0d6b58b
analysisStatus: analysis_failed · resultAttemptId: 8371777c-b23e-4b35-b46d-1e02a0d6b58b
- mentionJudgment: unknown
- recommendationJudgment: unknown
无法确认: Unexpected end of JSON input
[实际请求与原文证据](./public-evidence.json)
#### Attempt 8371777c-b23e-4b35-b46d-1e02a0d6b58b
analysis_failed · 时间: 2026-09-08T06:06:14.198Z · 首次 Attempt
requestedSearch: provider_native · used: true · executionMode: native
OpenRouter metadata confirmed provider-native web search.
错误: analysis_failed · Unexpected end of JSON input
finish_reason: stop
查看模型原始回答
{"analysisStatus":"completed","mentions":[{"name":"Pendo","domain":"pendo.io","recommendation":"positive","mentionQuote":"Pendo's analytics tool that instantly tracks what drives user engagement, retention, and product success—no manual event tagging needed.","recommendationQuote":"Pendo has been a huge win for us in this time when every dollar counts and efficiency is super important. It’s made collaboration more grounded in tangible objectives and results, and it provides metrics that define what ‘better’ really means.","firstMentionOffset":0,"firstRecommendationOffset":0,"firstMentionState":"unique","firstRecommendationState":"unique"},{"name":"Amplitude","domain":"amplitude.com","recommendation":"positive","mentionQuote":"Amplitude is a powerful product intelligence platform that will empower your product managers to self-serve and, in turn, allow you to make better and faster product decisions.","recommendationQuote":"Amplitude is a powerful product intelligence platform that will empower your product managers to self-serve and, in turn, allow you to make better and faster product decisions.","firstMentionOffset":0,"firstRecommendationOffset":0,"firstMentionState":"unique","firstRecommendationState":"unique"},{"name":"Mixpanel","domain":"mixpanel.com","recommendation":"positive","mentionQuote":"Mixpanel is a powerful product analytics tool that helps you convert, engage, and retain more users. Build funnels, see top user flows, create cohorts, and more with just a few clicks.","recommendationQuote":"Mixpanel is a powerful product analytics tool that helps you convert, engage, and retain more users. Build funnels, see top user flows, create cohorts, and more with just a few clicks.","firstMentionOffset":0,"firstRecommendationOffset":0,"firstMentionState":"unique","firstRecommendationState":"unique"},{"name":"PostHog","domain":"posthog.com","recommendation":"positive","mentionQuote":"PostHog is an open-source product analytics for developers. It helps engineers understand their product usage, automate events, and user data collection.","recommendationQuote":"PostHog is an open-source product analytics for developers. It helps engineers understand their product usage, automate events, and user data collection.","firstMentionOffset":0,"firstRecommendationOffset":0,"firstMentionState":"unique","firstRecommendationState":"unique"},{"name":"Heap","domain":"heap.io","recommendation":"positive","mentionQuote":"Heap is the only digital insights platform that gives you complete understanding of your customers’ digital journeys, so you can quickly improve conversion, retention, and customer delight.","recommendationQuote":"Heap is the only digital insights platform that gives you complete understanding of your customers’ digital journeys, so you can quickly improve conversion, retention, and customer delight.","firstMentionOffset":0,"firstRecommendationOffset":0,"firstMentionState":"unique","firstRecommendationState":"unique"},{"name":"LogRocket","domain":"logrocket.com","recommendation":"positive","mentionQuote":"LogRocket is a frontend monitoring and product analytics platform that combines high-fidelity session replay with technical telemetry to fix bugs and optimize UX.","recommendationQuote":"LogRocket is a frontend monitoring and product analytics platform that combines high-fidelity session replay with technical telemetry to fix bugs and optimize UX.","firstMentionOffset":0,"firstRecommendationOffset":0,"firstMentionState":"unique","firstRecommendationState":"unique"},{"name":"Whatfix","domain":"whatfix.com","recommendation":"positive","mentionQuote":"Whatfix Product Analytics goes beyond traditional dashboards by analyzing every user journey to reveal the ‘why’ behind the data. Pinpoint friction, fix issues faster, and deliver targeted in-app experiences that boost adoption and retention.","recommendationQuote":"Whatfix Product Analytics goes beyond traditional dashboards by analyzing every user journey to reveal the ‘why’ behind the data. Pinpoint friction, fix issues faster, and deliver targeted in-app experiences that boost adoption and retention.","firstMentionOffset":0,"firstRecommendationOffset":0,"firstMentionState":"unique","firstRecommendationState":"unique"},{"name":"Countly","domain":"count.ly","recommendation":"positive","mentionQuote":"Countly is an analytics platform to understand and enhance customer journeys in web, desktop and mobile applications. On-premise or in the Cloud.","recommendationQuote":"Countly is an analytics platform to understand and enhance customer journeys in web, desktop and mobile applications. On-premise or in the Cloud.","firstMentionOffset":
SHA-256: `4f79cefc23f91a03c719ac21b593d4f32d650fb324438739c5e022b3fd0b3576`
#### Provider 引用
此 Attempt 未归档此类来源。
#### 搜索检索结果
此 Attempt 未归档此类来源。
#### 回答中的链接(非 Provider 引用)
此 Attempt 未归档正文 URL。
### Feature Flags · openai/gpt-4o-mini
keywordId: watch-keyword-058bb884e2e3e12f02113bfd · runId: a99ba9b6-00fb-49e3-9f15-14a3faed91ae · probeId: 56894622-d030-4778-bf3c-5a9b70cca4e5
off · completed · firstAttemptId: d30e5a9f-4e39-46c4-8781-634e88b00262
analysisStatus: completed · resultAttemptId: d30e5a9f-4e39-46c4-8781-634e88b00262
- mentionJudgment: adjudicable
- recommendationJudgment: adjudicable
- Feature Flags: mentioned · mention: Feature Flags are a powerful tool for managing software features and deployments. · recommendation: — · attemptId: d30e5a9f-4e39-46c4-8781-634e88b00262
无法确认: —
[实际请求与原文证据](./public-evidence.json)
#### Attempt d30e5a9f-4e39-46c4-8781-634e88b00262
completed · 时间: 2026-09-08T06:07:08.722Z · 首次 Attempt
requestedSearch: off · used: false · executionMode: unverified
finish_reason: stop
查看模型原始回答
{"analysisStatus":"completed","mentions":[{"name":"Feature Flags","domain":null,"recommendation":"mentioned","mentionQuote":"Feature Flags are a powerful tool for managing software features and deployments.","recommendationQuote":null,"firstMentionOffset":0,"firstRecommendationOffset":null,"firstMentionState":"unique","firstRecommendationState":"none"}],"unknowns":[]}
SHA-256: `50f2c358be54909aed7f990b6d7d8b096d50c5c27806912a4769d684eb826cb2`
#### Provider 引用
此 Attempt 未归档此类来源。
#### 搜索检索结果
此 Attempt 未归档此类来源。
#### 回答中的链接(非 Provider 引用)
此 Attempt 未归档正文 URL。
### Product Analytics · openai/gpt-4o-mini
keywordId: watch-keyword-532e310d54beebe2356227d4 · runId: a99ba9b6-00fb-49e3-9f15-14a3faed91ae · probeId: 082d6c76-971c-4b96-abd9-10272232c88f
off · completed · firstAttemptId: 21572d72-609f-40ec-9162-fad31d65ceb3
analysisStatus: completed · resultAttemptId: 21572d72-609f-40ec-9162-fad31d65ceb3
- mentionJudgment: adjudicable
- recommendationJudgment: adjudicable
- Product Analytics: mentioned · mention: Product Analytics is a crucial aspect of understanding user behavior and improving product performance. · recommendation: — · attemptId: 21572d72-609f-40ec-9162-fad31d65ceb3
无法确认: —
[实际请求与原文证据](./public-evidence.json)
#### Attempt 21572d72-609f-40ec-9162-fad31d65ceb3
completed · 时间: 2026-09-08T06:07:13.141Z · 首次 Attempt
requestedSearch: off · used: false · executionMode: unverified
finish_reason: stop
查看模型原始回答
{"analysisStatus":"completed","mentions":[{"name":"Product Analytics","domain":null,"recommendation":"mentioned","mentionQuote":"Product Analytics is a crucial aspect of understanding user behavior and improving product performance.","recommendationQuote":null,"firstMentionOffset":0,"firstRecommendationOffset":null,"firstMentionState":"unique","firstRecommendationState":"none"}],"unknowns":[]}
SHA-256: `640b72a178649ee1457cab4da0a4176b38539500ae8d976f18de144f5e8da382`
#### Provider 引用
此 Attempt 未归档此类来源。
#### 搜索检索结果
此 Attempt 未归档此类来源。
#### 回答中的链接(非 Provider 引用)
此 Attempt 未归档正文 URL。
### Feature Flags · google/gemini-2.5-flash-lite
keywordId: watch-keyword-058bb884e2e3e12f02113bfd · runId: a99ba9b6-00fb-49e3-9f15-14a3faed91ae · probeId: ec46bc50-687c-43c0-a2a6-6743bf216885
off · completed · firstAttemptId: 42d9ffa7-f744-49fa-a869-c0e2e508720a
analysisStatus: completed · resultAttemptId: 42d9ffa7-f744-49fa-a869-c0e2e508720a
- mentionJudgment: adjudicable
- recommendationJudgment: adjudicable
- Feature Flags: mentioned · mention: Feature Flags · recommendation: — · attemptId: 42d9ffa7-f744-49fa-a869-c0e2e508720a
无法确认: —
[实际请求与原文证据](./public-evidence.json)
#### Attempt 42d9ffa7-f744-49fa-a869-c0e2e508720a
completed · 时间: 2026-09-08T06:07:02.778Z · 首次 Attempt
requestedSearch: off · used: false · executionMode: unverified
finish_reason: stop
查看模型原始回答
{
"analysisStatus": "completed",
"mentions": [
{
"name": "Feature Flags",
"domain": null,
"recommendation": "mentioned",
"mentionQuote": "Feature Flags",
"recommendationQuote": null,
"firstMentionOffset": 0,
"firstRecommendationOffset": null,
"firstMentionState": "unique",
"firstRecommendationState": "none"
}
],
"unknowns": []
}
SHA-256: `6d346060e85b092bba46e1684aa09c57db313bfc948da3d3e0e27ba0da25cf2e`
#### Provider 引用
此 Attempt 未归档此类来源。
#### 搜索检索结果
此 Attempt 未归档此类来源。
#### 回答中的链接(非 Provider 引用)
此 Attempt 未归档正文 URL。
### Product Analytics · google/gemini-2.5-flash-lite
keywordId: watch-keyword-532e310d54beebe2356227d4 · runId: a99ba9b6-00fb-49e3-9f15-14a3faed91ae · probeId: 9dfc6241-3c2b-437d-a8b0-9f0557bc501e
off · completed · firstAttemptId: 8a5eeec8-32e5-461f-8040-9c67d9c56507
analysisStatus: completed · resultAttemptId: 8a5eeec8-32e5-461f-8040-9c67d9c56507
- mentionJudgment: adjudicable
- recommendationJudgment: adjudicable
- Product Analytics: mentioned · mention: Product Analytics · recommendation: — · attemptId: 8a5eeec8-32e5-461f-8040-9c67d9c56507
无法确认: —
[实际请求与原文证据](./public-evidence.json)
#### Attempt 8a5eeec8-32e5-461f-8040-9c67d9c56507
completed · 时间: 2026-09-08T06:07:09.628Z · 首次 Attempt
requestedSearch: off · used: false · executionMode: unverified
finish_reason: stop
查看模型原始回答
{
"analysisStatus": "completed",
"mentions": [
{
"name": "Product Analytics",
"domain": null,
"recommendation": "mentioned",
"mentionQuote": "Product Analytics",
"recommendationQuote": null,
"firstMentionOffset": 0,
"firstRecommendationOffset": null,
"firstMentionState": "unique",
"firstRecommendationState": "none"
}
],
"unknowns": []
}
SHA-256: `848c0ad989bb26d0ecb5bfc9f319d02a19ae0e817ad211a7278db240872fd70b`
#### Provider 引用
此 Attempt 未归档此类来源。
#### 搜索检索结果
此 Attempt 未归档此类来源。
#### 回答中的链接(非 Provider 引用)
此 Attempt 未归档正文 URL。
### Feature Flags · openai/gpt-4.1-mini
keywordId: watch-keyword-058bb884e2e3e12f02113bfd · runId: a99ba9b6-00fb-49e3-9f15-14a3faed91ae · probeId: 87e02919-a3e5-42ce-b2da-bae153456de0
provider_native · completed · firstAttemptId: 58019048-c5a6-4b67-af85-c23f0d7450b0
analysisStatus: completed · resultAttemptId: 58019048-c5a6-4b67-af85-c23f0d7450b0
- mentionJudgment: adjudicable
- recommendationJudgment: adjudicable
- Feature Flags: mentioned · mention: Feature flags are a runtime switch in code that turns a piece of functionality on or off without redeploying. ([glossary.deployment.to](https://glossary.deployment.to/feature-flag/?utm_source=openai)) · recommendation: — · attemptId: 58019048-c5a6-4b67-af85-c23f0d7450b0
无法确认: —
[实际请求与原文证据](./public-evidence.json)
#### Attempt 58019048-c5a6-4b67-af85-c23f0d7450b0
completed · 时间: 2026-09-08T06:07:06.429Z · 首次 Attempt
requestedSearch: provider_native · used: true · executionMode: native
OpenRouter metadata confirmed provider-native web search.
finish_reason: stop
查看模型原始回答
{"analysisStatus":"completed","mentions":[{"name":"Feature Flags","domain":null,"recommendation":"mentioned","mentionQuote":"Feature flags are a runtime switch in code that turns a piece of functionality on or off without redeploying. ([glossary.deployment.to](https://glossary.deployment.to/feature-flag/?utm_source=openai))","recommendationQuote":null,"firstMentionOffset":0,"firstRecommendationOffset":0,"firstMentionState":"unique","firstRecommendationState":"none"}],"unknowns":[]}
SHA-256: `9ff4986b91f6ddf9cf910513b7ec8819746ccc826ff48eb30de7e25d027f0462`
#### Provider 引用
- What is a feature flag? Meaning + example - deployment.to: [https://glossary.deployment.to/feature-flag/?utm_source=openai]() · `choices[0].message.annotations[0].url_citation.url`
#### 搜索检索结果
此 Attempt 未归档此类来源。
#### 回答中的链接(非 Provider 引用)
- [https://glossary.deployment.to/feature-flag/?utm_source=openai]()
### Product Analytics · openai/gpt-4.1-mini
keywordId: watch-keyword-532e310d54beebe2356227d4 · runId: a99ba9b6-00fb-49e3-9f15-14a3faed91ae · probeId: 96ab1562-739a-4dac-8b68-982d788e10c9
provider_native · failed · firstAttemptId: e2716c84-de3a-491a-83cb-6f63765dd862
analysisStatus: analysis_failed · resultAttemptId: e2716c84-de3a-491a-83cb-6f63765dd862
- mentionJudgment: unknown
- recommendationJudgment: unknown
无法确认: Unterminated string in JSON at position 4535 (line 1 column 4536)
[实际请求与原文证据](./public-evidence.json)
#### Attempt e2716c84-de3a-491a-83cb-6f63765dd862
analysis_failed · 时间: 2026-09-08T06:07:12.056Z · 首次 Attempt
requestedSearch: provider_native · used: true · executionMode: native
OpenRouter metadata confirmed provider-native web search.
错误: analysis_failed · Unterminated string in JSON at position 4535 (line 1 column 4536)
finish_reason: stop
查看模型原始回答
{"analysisStatus":"completed","mentions":[{"name":"Pendo","domain":"pendo.io","recommendation":"positive","mentionQuote":"Pendo's analytics tool that instantly tracks what drives user engagement, retention, and product success—no manual event tagging needed.","recommendationQuote":"Pendo has been a huge win for us in this time when every dollar counts and efficiency is super important. It’s made collaboration more grounded in tangible objectives and results, and it provides metrics that define what ‘better’ really means.","firstMentionOffset":0,"firstRecommendationOffset":0,"firstMentionState":"unique","firstRecommendationState":"unique"},{"name":"Google Analytics","domain":"analytics.google.com","recommendation":"positive","mentionQuote":"Google Analytics is our overall pick for product analytics in 2026.","recommendationQuote":"Google Analytics is our overall pick for product analytics in 2026.","firstMentionOffset":0,"firstRecommendationOffset":0,"firstMentionState":"unique","firstRecommendationState":"unique"},{"name":"Firebase Analytics","domain":"firebase.google.com/products/analytics","recommendation":"positive","mentionQuote":"Firebase Analytics is our free pick for product analytics in 2026.","recommendationQuote":"Firebase Analytics is our free pick for product analytics in 2026.","firstMentionOffset":0,"firstRecommendationOffset":0,"firstMentionState":"unique","firstRecommendationState":"unique"},{"name":"Amplitude","domain":"amplitude.com","recommendation":"positive","mentionQuote":"Amplitude is a powerful product intelligence platform that will empower your product managers to self-serve and, in turn, allow you to make better and faster product decisions.","recommendationQuote":"Amplitude is a powerful product intelligence platform that will empower your product managers to self-serve and, in turn, allow you to make better and faster product decisions.","firstMentionOffset":0,"firstRecommendationOffset":0,"firstMentionState":"unique","firstRecommendationState":"unique"},{"name":"Mixpanel","domain":"mixpanel.com","recommendation":"positive","mentionQuote":"Mixpanel is our solid pick for product analytics in 2026.","recommendationQuote":"Mixpanel is our solid pick for product analytics in 2026.","firstMentionOffset":0,"firstRecommendationOffset":0,"firstMentionState":"unique","firstRecommendationState":"unique"},{"name":"PostHog","domain":"posthog.com","recommendation":"positive","mentionQuote":"PostHog is our solid pick for product analytics in 2026.","recommendationQuote":"PostHog is our solid pick for product analytics in 2026.","firstMentionOffset":0,"firstRecommendationOffset":0,"firstMentionState":"unique","firstRecommendationState":"unique"},{"name":"LogRocket","domain":"logrocket.com","recommendation":"positive","mentionQuote":"LogRocket is our most affordable pick for product analytics in 2026.","recommendationQuote":"LogRocket is our most affordable pick for product analytics in 2026.","firstMentionOffset":0,"firstRecommendationOffset":0,"firstMentionState":"unique","firstRecommendationState":"unique"},{"name":"ChurnZero","domain":"churnzero.com","recommendation":"positive","mentionQuote":"ChurnZero is our community favorite pick for product analytics in 2026.","recommendationQuote":"ChurnZero is our community favorite pick for product analytics in 2026.","firstMentionOffset":0,"firstRecommendationOffset":0,"firstMentionState":"unique","firstRecommendationState":"unique"},{"name":"mParticle","domain":"www.mparticle.com","recommendation":"positive","mentionQuote":"mParticle is a product analytics platform that connects directly to your data warehouse to provide actionable insights across the entire customer journey.","recommendationQuote":"mParticle is a product analytics platform that connects directly to your data warehouse to provide actionable insights across the entire customer journey.","firstMentionOffset":0,"firstRecommendationOffset":0,"firstMentionState":"unique","firstRecommendationState":"unique"},{"name":"Heap","domain":"heap.io","recommendation":"positive","mentionQuote":"Heap is the only digital insights platform that gives you complete understanding of your customers’ digital journeys, so you can quickly improve conversion, retention, and customer delight.","recommendationQuote":"Heap is the only digital insights platform that gives you complete understanding of your customers’ digital journeys, so you can quickly improve conversion, retention, and customer delight.","firstMentionOffset":0,"firstRecommendationOffset
SHA-256: `7407f11da9400f5b8d3a07c0e8de39d938ade3b4f3b1fca619ecc9b7c01271fe`
#### Provider 引用
此 Attempt 未归档此类来源。
#### 搜索检索结果
此 Attempt 未归档此类来源。
#### 回答中的链接(非 Provider 引用)
此 Attempt 未归档正文 URL。
## 重复观察
3 次归档测量运行;数量不表示全部成功。只比较相同模型、语言、搜索与指纹条件。短期重复不代表长期趋势或因果效果。
- Run 942c0129-c245-4a00-9810-3f8bcde4d518: partial
- Run 71466c93-74d0-4807-9a42-64bdb7cabc78: partial
- Run a99ba9b6-00fb-49e3-9f15-14a3faed91ae: partial
- D 4924864f-0dab-4595-977c-afeeaecfb765 · openai/gpt-4.1-mini · domainRecognition: recognized · analysisStatus: recognized · firstAttemptId: e1488831-b6ea-4dd3-b6ee-588afda4b4ad · resultAttemptId: e1488831-b6ea-4dd3-b6ee-588afda4b4ad
- D 374fe0ec-52f1-4f41-bada-a0f09c37409a · openai/gpt-4o-mini · domainRecognition: recognized · analysisStatus: recognized · firstAttemptId: 4cf7045c-8e37-49fb-94c0-d52bd7186109 · resultAttemptId: 4cf7045c-8e37-49fb-94c0-d52bd7186109
- D 5bc305f1-028b-4092-8673-26347ae5d9b5 · google/gemini-2.5-flash-lite · domainRecognition: recognized · analysisStatus: recognized · firstAttemptId: a1dd1769-1aa2-472d-9b54-27864bfbedca · resultAttemptId: a1dd1769-1aa2-472d-9b54-27864bfbedca
- D 048ceb56-7933-458a-9610-69c8f9932ab7 · openai/gpt-4o-mini · domainRecognition: recognized · analysisStatus: recognized · firstAttemptId: 2f256c93-88e3-403e-a478-00e0b04fb08a · resultAttemptId: 2f256c93-88e3-403e-a478-00e0b04fb08a
- D 98933c48-b6e8-40af-a28c-da5315557ff0 · google/gemini-2.5-flash-lite · domainRecognition: recognized · analysisStatus: recognized · firstAttemptId: d7824db9-8fea-4a4f-8a48-1cb063f8887c · resultAttemptId: d7824db9-8fea-4a4f-8a48-1cb063f8887c
- D 5361fa95-e877-4424-8f25-cf4cb06786cc · openai/gpt-4.1-mini · domainRecognition: recognized · analysisStatus: recognized · firstAttemptId: af1dd88f-7e2c-4fb8-9429-8aa80cbcae86 · resultAttemptId: af1dd88f-7e2c-4fb8-9429-8aa80cbcae86
- D 5d447242-3122-42e7-9bfa-2f9c5363a8bd · openai/gpt-4o-mini · domainRecognition: recognized · analysisStatus: recognized · firstAttemptId: c5d5f99f-0ed7-4a62-b173-19e09ad9d851 · resultAttemptId: c5d5f99f-0ed7-4a62-b173-19e09ad9d851
- D 4bd9b598-3650-45ad-a669-03dbe6b55df4 · google/gemini-2.5-flash-lite · domainRecognition: recognized · analysisStatus: recognized · firstAttemptId: d45926fe-1290-4bdf-9b7c-243746ce3f31 · resultAttemptId: d45926fe-1290-4bdf-9b7c-243746ce3f31
- D c04a2dec-2c22-4b65-9268-220870e475ca · openai/gpt-4.1-mini · domainRecognition: recognized · analysisStatus: recognized · firstAttemptId: 88abb312-90da-4a0b-9782-8dc864c08fa3 · resultAttemptId: 88abb312-90da-4a0b-9782-8dc864c08fa3
### 调度归档
收集时任务状态(历史快照): active
最终任务归档状态: paused · nextRunAt: — · updatedAt: 2026-09-08T06:07:24.138Z
product-data/projects/b4b79062-5865-4b82-8d3a-dd16d5dc68cd/schedules/tasks/f711adc1-1c20-4f71-981e-5af0e7d2c475.json · SHA-256: `ded6cd12c3e62866497b38e7c8019cc63629c173bf45890f13ee2347f450fbea`
- Occurrence c00d21b0-076d-4ac4-a0bd-941670509b57: completed · reason: run_partial_or_failed · runId: a99ba9b6-00fb-49e3-9f15-14a3faed91ae · scheduledFor: 2026-09-08T06:07:00.000Z
Occurrence completed 表示调度记录结束;关联 Run 的 partial/failed 仍是部分失败或失败。
## 产品截图

R04 · posthog.com · D · 3 条归档尝试 · openai/gpt-4o-mini(off)、google/gemini-2.5-flash-lite(off)、openai/gpt-4.1-mini(provider_native) · 2026-09-08T06:05:22.972Z 至 2026-09-08T06:05:22.972Z。保留原始失败状态。
截图时间: 2026-09-08T07:18:58.029Z.

R04 · posthog.com · D/K · 27 条归档尝试 · openai/gpt-4o-mini(off)、google/gemini-2.5-flash-lite(off)、openai/gpt-4.1-mini(provider_native) · 2026-09-08T06:05:33.040Z 至 2026-09-08T06:07:12.056Z。保留原始失败状态。
截图时间: 2026-09-08T07:18:58.461Z.

R04 · posthog.com · D/K · 27 条归档尝试 · openai/gpt-4o-mini(off)、google/gemini-2.5-flash-lite(off)、openai/gpt-4.1-mini(provider_native) · 2026-09-08T06:05:33.040Z 至 2026-09-08T06:07:12.056Z。保留原始失败状态。
截图时间: 2026-09-08T07:19:00.202Z.
## 复核与重新测量
[案例配置](./case.json) · [证据索引](./evidence-index.json) · [公开证据包](./public-evidence.json)
SHA-256: `ea36fddd9c9cea5ec76b3839ec6e7662453d68ee9a4f62cc5d96c3d7ee8b4174`
历史案例费用(非本轮文档费用): USD 0.14683270 · 30 次调用 · 106147 Token.
[导出源码](../../../examples/lib/export.mjs) · [候选版本与未发布状态](../../../docs/releases/v0.2.0-rc.1.md)
```bash
npm run examples:plan -- --case R04
npm run examples:replay -- --case R04 --evidence examples/cases/R04/public-evidence.json
```
重新测量会收费:必须使用独立产品服务、冻结计划和显式 live 授权。阅读或导出不调用模型。
## 限制与披露
仅作公开产品观察;收录不表示合作或背书关系。
模型描述不是已核实的市场事实。来源存在不证明页面内容正确,也不说明为何推荐。未知、缺失、解析失败与请求失败保持区分。可据此调查业务误述或来源内容,但不能断言差距由某次优化造成。
- Run 942c0129-c245-4a00-9810-3f8bcde4d518: partial
- Run 71466c93-74d0-4807-9a42-64bdb7cabc78: partial
- Run a99ba9b6-00fb-49e3-9f15-14a3faed91ae: partial
- Probe d85e30b0-0dd7-46d9-900c-7c0cb22273e4: failed; first attempt analysis_failed
- Probe d85e30b0-0dd7-46d9-900c-7c0cb22273e4: missing or failed analysis
- Probe 2713f3dd-9000-4378-8f92-36c029d3dae3: failed; first attempt analysis_failed
- Probe 2713f3dd-9000-4378-8f92-36c029d3dae3: missing or failed analysis
- Probe 96ab1562-739a-4dac-8b68-982d788e10c9: failed; first attempt analysis_failed
- Probe 96ab1562-739a-4dac-8b68-982d788e10c9: missing or failed analysis
- Attempt 829eb19d-0fb0-4d0d-9055-1a7e62188ffd: analysis_failed; Unterminated string in JSON at position 4368 (line 1 column 4369)
- Attempt 8371777c-b23e-4b35-b46d-1e02a0d6b58b: analysis_failed; Unexpected end of JSON input
- Attempt e2716c84-de3a-491a-83cb-6f63765dd862: analysis_failed; Unterminated string in JSON at position 4535 (line 1 column 4536)
- Occurrence c00d21b0-076d-4ac4-a0bd-941670509b57: completed; run_partial_or_failed