indiebiz-evidence
chilayt/indiebiz-evidence
Cross-check collected data across sources, with provenance, before trusting it. Zero-dependency MCP.
- tools
- 15
- transport
- STDIO
- last updated
- Nov 13, 2026
Tools
15 toolsPublic tool metadata for what this MCP can expose to an agent.
list_sources已导入的源、时间范围、实体/字段数（先调这个）
已导入的源、时间范围、实体/字段数（先调这个）
No parameter schema in public metadata yet.
audit_field_coverage采集健康度：逐字段覆盖率、absent_on_latest 判定、schema_events、断档节奏
采集健康度：逐字段覆盖率、absent_on_latest 判定、schema_events、断档节奏
No parameter schema in public metadata yet.
search_records任意字段筛选 + 排序 + 分页；缺席字段永不满足比较
任意字段筛选 + 排序 + 分页；缺席字段永不满足比较
No parameter schema in public metadata yet.
get_source_entity单实体某日全字段快照
单实体某日全字段快照
No parameter schema in public metadata yet.
query_series单字段全时序（缺口 = 缺席，不是 0）
单字段全时序（缺口 = 缺席，不是 0）
No parameter schema in public metadata yet.
get_source_evidence_chain单实体全字段跨日审计链 + 逐日 raw_ref
单实体全字段跨日审计链 + 逐日 raw_ref
No parameter schema in public metadata yet.
diff_snapshots两次抓取之间变了什么（新增/消失/逐字段），与领域无关
两次抓取之间变了什么（新增/消失/逐字段），与领域无关
No parameter schema in public metadata yet.
list_entities源内实体清单（首末次出现 / 出现的抓取日数 / 事实数）
源内实体清单（首末次出现 / 出现的抓取日数 / 事实数）
No parameter schema in public metadata yet.
find_duplicate_entities源内合并候选，带逐对证据
源内合并候选，带逐对证据
No parameter schema in public metadata yet.
resolve_entities跨源找同一个东西的各种写法，只给候选，不自动合并
跨源找同一个东西的各种写法，只给候选，不自动合并
No parameter schema in public metadata yet.
cross_check跨源核对：同一实体在所有来源的值摆一起，判一致/打架/孤证，各带出处
跨源核对：同一实体在所有来源的值摆一起，判一致/打架/孤证，各带出处
No parameter schema in public metadata yet.
propose_entity_links跨源对齐候选：同名/近名实体 + 逐字段一致/冲突证据；绝不自动连
跨源对齐候选：同名/近名实体 + 逐字段一致/冲突证据；绝不自动连
No parameter schema in public metadata yet.
confirm_entity_link记录人工确认的跨源对齐（source+entity→canonical_id，存谁/何时/为何）；唯一建链方式
记录人工确认的跨源对齐（source+entity→canonical_id，存谁/何时/为何）；唯一建链方式
No parameter schema in public metadata yet.
list_entity_links已确认的对齐，按 canonical_id 分组
已确认的对齐，按 canonical_id 分组
No parameter schema in public metadata yet.
list_source_anomalies检测到的异常：数值摆动、累计回退、字段/实体消失、文本漂移、源改版、抓取断档
检测到的异常：数值摆动、累计回退、字段/实体消失、文本漂移、源改版、抓取断档
No parameter schema in public metadata yet.
Sponsors
Promote your product hereReach 125k+ visitors monthlyPromote your product hereReach 125k+ visitors monthly
Agent CaffeineSponsoredClose your lid, come back to finished work
SnugSponsoredThe clipboard manager for the AI eraPromote your product hereReach 125k+ visitors monthly
WaxSponsoredBlock anything with supervision, without erasing your iPhonePromote your product hereReach 125k+ visitors monthly
knobsSponsoredSee your site on any phone, any day, in any language
Agent CaffeineSponsoredClose your lid, come back to finished workSnugSponsoredThe clipboard manager for the AI eraPromote your product hereReach 125k+ visitors monthlyWaxSponsoredBlock anything with supervision, without erasing your iPhonePromote your product hereReach 125k+ visitors monthlyknobsSponsoredSee your site on any phone, any day, in any languageReadme
indiebiz-evidence
English — You collect information from several sources. Before you present a number as "the answer", you need to know: do the sources agree? Can you trace it back to the raw file? Is it even still being reported? indiebiz-evidence is a local, zero-dependency evidence layer that cross-checks what you collected — agreement across sources, provenance, and anomalies — so you verify before you conclude. It does not collect for you; it helps you verify what you collected — equal values are not proof of truth. Point it at dated CSV/JSON exports and query it from the CLI or a zero-dependency MCP server your agent can call.
你从好几个来源收集信息。但在把一个数字当成"答案"给出去之前，你得先知道：这几个来源说的 一致吗？能不能回溯到原始文件？它现在还在被报告吗？ indiebiz-evidence 是一个本地、零依赖的 证据层，把你收集到的东西交叉复核——多源是否一致、出处可回溯、异常先声明——让你先核实， 再下结论。它不替你采集，它帮你核实采到的东西——值相同不等于事实已核实。指向按日期命名的 CSV/JSON 导出即可，CLI 或 零依赖 MCP Server 供你的 agent 调用。
它解决的不是"抓不到"，是"没复核就当成正确答案"
做信息收集的人最容易翻的车，不是抓取失败，而是复核不完全就把一个值当成了事实：
Details
- Repository
- github.com/chilayt/indiebiz-evidence
- Category
- Data & Analytics
- Package
indiebiz-evidence- Transport
- STDIO
- Registry
- MCP registry, active
