{"number":"03","slug":"evidence-provenance-standard","filename":"03-evidence-provenance-standard.md","title":"Evidence and Provenance Standard","bytes":10252,"lines":228,"sha256":"d67fb457e533a5b205efaa200cda769ae21be6e1556ce7f98bab7708812e7474","sections":14,"tables":3,"code_blocks":5,"href":"/v1/docs/evidence-provenance-standard","outline":[{"anchor":"1-evidence-states","level":2,"title":"1. Evidence states","line":3,"tables":["03.1-evidence-states.t1"],"code":[]},{"anchor":"2-evidence-record","level":2,"title":"2. Evidence record","line":18,"tables":[],"code":["03.2-evidence-record.c1"]},{"anchor":"3-claim-record","level":2,"title":"3. Claim record","line":51,"tables":[],"code":["03.3-claim-record.c1"]},{"anchor":"4-confidence-model","level":2,"title":"4. Confidence model","line":76,"tables":[],"code":["03.4-confidence-model.c1"]},{"anchor":"5-triangulation-rules","level":2,"title":"5. Triangulation rules","line":106,"tables":[],"code":[]},{"anchor":"6-temporal-rules","level":2,"title":"6. Temporal rules","line":118,"tables":[],"code":[]},{"anchor":"7-sentiment-evidence","level":2,"title":"7. Sentiment evidence","line":126,"tables":[],"code":["03.7-sentiment-evidence.c1"]},{"anchor":"8-review-and-complaint-evidence","level":2,"title":"8. Review and complaint evidence","line":148,"tables":[],"code":[]},{"anchor":"9-security-evidence","level":2,"title":"9. Security evidence","line":156,"tables":["03.9-security-evidence.t1"],"code":[]},{"anchor":"10-financial-evidence","level":2,"title":"10. Financial evidence","line":170,"tables":[],"code":["03.10-financial-evidence.c1"]},{"anchor":"11-contradiction-protocol","level":2,"title":"11. Contradiction protocol","line":184,"tables":[],"code":[]},{"anchor":"12-reproducibility-packet","level":2,"title":"12. Reproducibility packet","line":195,"tables":[],"code":[]},{"anchor":"13-evidence-use-matrix","level":2,"title":"13. Evidence-use matrix","line":211,"tables":["03.13-evidence-use-matrix.t1"],"code":[]}],"blocks":[{"type":"heading","level":1,"text":"Evidence and Provenance Standard","anchor":"evidence-and-provenance-standard","line":1},{"type":"heading","level":2,"text":"1. Evidence states","anchor":"1-evidence-states","line":3},{"type":"table","headers":["Level","Name","Minimum condition","Allowed use"],"align":["left","left","left","left"],"rows":[["E0","No evidence","no preserved support","backlog only"],["E1","Lead","snippet, unverified mention, weakly resolved source","research lead; not report fact"],["E2","Single-source support","accessible source with extract and time, but secondary, user-generated, or not corroborated","bounded observation with caveat"],["E3","Current official reference snapshot","current official page, filing, registry, API, policy, repository, or first-party platform record preserved with provenance","direct claim about what source shows"],["E4","Corroborated official evidence","E3 plus independent primary/authoritative corroboration or an authorized reproducible local measurement/registry","material finding and model input"],["E5","Confirmed/audited evidence","E4 support plus audited/reconciled target data or a valid controlled test that directly addresses the scoped claim","high-confidence decision input"]],"line":5,"id":"03.1-evidence-states.t1"},{"type":"paragraph","text":"E3/E4 preserves the user's established distinction: official current documentation is E3; official current documentation corroborated by an installed registry or independent reproducible evidence is E4.","line":14},{"type":"paragraph","text":"Evidence level is not truth probability. An official source can be outdated or self-interested; a customer review can accurately describe one experience. Target-confirmed but unaudited data does not automatically outrank corroborated evidence; classify its provenance and validate it under the appropriate E3/E4 level. Claim scope must match what the evidence proves.","line":16},{"type":"heading","level":2,"text":"2. Evidence record","anchor":"2-evidence-record","line":18},{"type":"code","lang":"yaml","content":"source_id: SRC-000001\ncanonical_url: \"\"\nsource_type: official|regulator|filing|registry|court|academic|platform|news|trade|review|forum|social|podcast|advertisement|archive|repository|dataset|measurement|other\npublisher: \"\"\ntitle: \"\"\nauthor: \"\"\npublished_at: \"\"\nobserved_at: \"\"\nretrieved_at: \"\"\nvalid_time_start: \"\"\nvalid_time_end: \"\"\njurisdiction: \"\"\naccess_class: public|authorized|restricted|unknown\ncollection_method: browser|official_api|rss|download|manual|passive_service\nquery_or_request: \"\"\nextract: \"\"\nstructured_facts: {}\ncontent_hash: \"\"\narchive_pointer: \"\"\nterms_notes: \"\"\npersonal_data_class: none|business_contact|personal|sensitive\nretention_rule: \"\"\nevidence_level: E0|E1|E2|E3|E4|E5\nreviewer: \"\"","line":20,"id":"03.2-evidence-record.c1"},{"type":"paragraph","text":"Do not store full copyrighted articles when a short extract plus URL and structured facts will do. Never store passwords, tokens, private keys, breach dumps, or unnecessary personal data.","line":47},{"type":"paragraph","text":"`source_type` is an extensible controlled vocabulary. New values require a registry definition and mapping; do not force a source into a misleading existing class.","line":49},{"type":"heading","level":2,"text":"3. Claim record","anchor":"3-claim-record","line":51},{"type":"code","lang":"yaml","content":"claim_id: CLM-000001\nclaim_text: \"\"\nsubject_entity_id: \"\"\npredicate: \"\"\nobject_or_value: \"\"\nclaim_state: O|C|I|H|U|X\nscope:\n  time: \"\"\n  geography: \"\"\n  product: \"\"\nevidence_links:\n  supports: []\n  contradicts: []\ncalculation_id: \"\"\nconfidence_band: low|moderate|high|very-high\nconfidence_score_internal: 0-100\nmateriality: low|medium|high|critical\nlast_reviewed_at: \"\"\nreview_status: draft|verified|challenged|retracted\nlimitations: []","line":53,"id":"03.3-claim-record.c1"},{"type":"heading","level":2,"text":"4. Confidence model","anchor":"4-confidence-model","line":76},{"type":"paragraph","text":"Use confidence to prioritize review, not to manufacture certainty.","line":78},{"type":"code","lang":"text","content":"Raw confidence =\n  0.25 SourceAuthority\n+ 0.20 Directness\n+ 0.15 Independence\n+ 0.15 Recency\n+ 0.15 Reproducibility\n+ 0.10 ScopeCompleteness\n\nAdjusted confidence = 100 × Raw confidence × ResolutionFactor × ContradictionFactor","line":80,"id":"03.4-confidence-model.c1"},{"type":"paragraph","text":"Score each component from 0 to 1. Use:","line":92},{"type":"list","ordered":false,"checklist":false,"items":[{"text":"`ResolutionFactor`: 1.0 exact entity; 0.75 probable; 0.4 ambiguous; 0 if unresolved.","depth":0,"checked":null},{"text":"`ContradictionFactor`: 1.0 none found after a real check; 0.7 unresolved conflict; 0.3 strong conflict.","depth":0,"checked":null}],"line":94},{"type":"paragraph","text":"Bands:","line":97},{"type":"list","ordered":false,"checklist":false,"items":[{"text":"0–39: low","depth":0,"checked":null},{"text":"40–64: moderate","depth":0,"checked":null},{"text":"65–84: high","depth":0,"checked":null},{"text":"85–100: very high","depth":0,"checked":null}],"line":99},{"type":"paragraph","text":"Only E3+ direct observations can normally reach high confidence. Inferences should rarely be labeled very high even when their inputs are strong.","line":104},{"type":"heading","level":2,"text":"5. Triangulation rules","anchor":"5-triangulation-rules","line":106},{"type":"paragraph","text":"Triangulation requires source independence, not three pages repeating one press release.","line":108},{"type":"paragraph","text":"For a material claim seek:","line":110},{"type":"list","ordered":true,"checklist":false,"items":[{"text":"one primary or official source;","depth":0,"checked":null},{"text":"one independent source or authorized measurement;","depth":0,"checked":null},{"text":"one temporal comparison or disconfirming check.","depth":0,"checked":null}],"line":112},{"type":"paragraph","text":"Record syndication and common-origin relationships. Treat reposts, copied reviews, mirrored databases, press-release pickups, and shared datasets as one origin.","line":116},{"type":"heading","level":2,"text":"6. Temporal rules","anchor":"6-temporal-rules","line":118},{"type":"list","ordered":false,"checklist":false,"items":[{"text":"Store **event time**, **publication time**, **valid time**, and **retrieval time** separately.","depth":0,"checked":null},{"text":"A current page can describe an old event; a recent article may repeat an obsolete fact.","depth":0,"checked":null},{"text":"For web changes, preserve the before/after source and first-seen/last-confirmed times.","depth":0,"checked":null},{"text":"A missing historical snapshot means `Unknown`, not “the page did not exist.”","depth":0,"checked":null},{"text":"Compare equivalent time windows; state partial-period data.","depth":0,"checked":null}],"line":120},{"type":"heading","level":2,"text":"7. Sentiment evidence","anchor":"7-sentiment-evidence","line":126},{"type":"paragraph","text":"Sentiment is a model output, not an observed emotion.","line":128},{"type":"paragraph","text":"Required fields:","line":130},{"type":"code","lang":"yaml","content":"text_unit_id: \"\"\nsource_id: \"\"\nlanguage: \"\"\ntopic: \"\"\nstance_target: \"\"\nsentiment_label: positive|neutral|negative|mixed|uncertain\nmodel_or_coding_method: \"\"\nmodel_version: \"\"\nhuman_validated_sample: \"\"\nsampling_frame: \"\"\nplatform_bias_notes: \"\"","line":132,"id":"03.7-sentiment-evidence.c1"},{"type":"paragraph","text":"Never report “customers are 40% negative” when the sample is actually “40% of collected comments were classified as negative.” Report the denominator, window, sources, sampling limits, and validation accuracy.","line":146},{"type":"heading","level":2,"text":"8. Review and complaint evidence","anchor":"8-review-and-complaint-evidence","line":148},{"type":"list","ordered":false,"checklist":false,"items":[{"text":"Separate star rating from text theme, recency, location, product, and company response.","depth":0,"checked":null},{"text":"Detect likely duplicates; do not remove adverse records merely because they are inconvenient.","depth":0,"checked":null},{"text":"Do not infer fraud, identity, or authenticity without platform verification or strong evidence.","depth":0,"checked":null},{"text":"Report complaint prevalence within the collected sample, not the customer population.","depth":0,"checked":null},{"text":"A BBB complaint or forum post is an allegation/experience report, not an adjudicated fact.","depth":0,"checked":null}],"line":150},{"type":"heading","level":2,"text":"9. Security evidence","anchor":"9-security-evidence","line":156},{"type":"paragraph","text":"Security observations require especially narrow language:","line":158},{"type":"table","headers":["Unsafe claim","Evidence-safe version"],"align":["left","left"],"rows":[["“They were breached”","“The domain appears in an authorized breach-notification result as of [date]; scope and current risk are unconfirmed.”"],["“They have no MFA”","“MFA status is not publicly verifiable.”"],["“This version is vulnerable”","“A passive fingerprint suggests version X; the version and exploitability were not validated.”"],["“Their endpoint is exposed”","“A passive index listed service Y at time Z; no active connection was attempted.”"],["“Employee credentials are on the dark web”","Do not make or store this claim from unverified dumps. Use an authorized domain-monitoring process."]],"line":160,"id":"03.9-security-evidence.t1"},{"type":"paragraph","text":"Do not put sensitive technical details in sales outreach. Route privately to a published security contact with minimal proof and no exploitation.","line":168},{"type":"heading","level":2,"text":"10. Financial evidence","anchor":"10-financial-evidence","line":170},{"type":"paragraph","text":"Every financial number is one of:","line":172},{"type":"code","lang":"text","content":"R = Reported by target/filing\nM = Measured from authorized data\nB = External benchmark\nA = Explicit assumption\nC = Calculated","line":174,"id":"03.10-financial-evidence.c1"},{"type":"paragraph","text":"Show the tag next to each input. A model made entirely from B/A inputs is a scenario, never an estimated loss stated as fact.","line":182},{"type":"heading","level":2,"text":"11. Contradiction protocol","anchor":"11-contradiction-protocol","line":184},{"type":"paragraph","text":"For each high-materiality claim:","line":186},{"type":"list","ordered":true,"checklist":false,"items":[{"text":"write the strongest plausible alternative;","depth":0,"checked":null},{"text":"search for disconfirming evidence;","depth":0,"checked":null},{"text":"compare source authority, scope, and time;","depth":0,"checked":null},{"text":"preserve both evidence paths;","depth":0,"checked":null},{"text":"narrow, downgrade, mark contradicted, or retract the claim;","depth":0,"checked":null},{"text":"state what new evidence would resolve it.","depth":0,"checked":null}],"line":188},{"type":"heading","level":2,"text":"12. Reproducibility packet","anchor":"12-reproducibility-packet","line":195},{"type":"paragraph","text":"Each audit should retain:","line":197},{"type":"list","ordered":false,"checklist":false,"items":[{"text":"scope contract and registry version;","depth":0,"checked":null},{"text":"tool/source versions where relevant;","depth":0,"checked":null},{"text":"queries or API request templates without secrets;","depth":0,"checked":null},{"text":"retrieval timestamps;","depth":0,"checked":null},{"text":"normalized source/claim ledgers;","depth":0,"checked":null},{"text":"calculations and assumptions;","depth":0,"checked":null},{"text":"decision/elimination log;","depth":0,"checked":null},{"text":"human-review record;","depth":0,"checked":null},{"text":"limitations and failed/blocked retrievals.","depth":0,"checked":null}],"line":199},{"type":"paragraph","text":"API keys, cookies, tokens, and credentials never enter the packet.","line":209},{"type":"heading","level":2,"text":"13. Evidence-use matrix","anchor":"13-evidence-use-matrix","line":211},{"type":"paragraph","text":"Evidence level and claim type jointly determine use. “Two sources” never substitutes automatically for an official source, and an official page is not required for every reproducible corpus calculation.","line":213},{"type":"table","headers":["Claim type","Report use","Outreach use"],"align":["left","left","left"],"rows":[["Official/company fact `[O]`","Current E3; E4/E5 for material decisions","Current E3+ with exact scope and role relevance"],["Reproducible public-corpus measure `[C]`","Material source items E2+ from at least two independent origins; declared query, window, denominator, dedupe, sampling limits, and reproducible calculation","Allowed as a calculated sample observation after human validation and defamation/privacy review; wording must describe the collected sample, not the population"],["Deterministic historical calculation `[C][DETERMINISTIC_CALCULATION]`","Reconciled inputs and formula; input evidence governs confidence","Allowed when inputs and scope are current/relevant; never imply causality"],["Financial projection `[C][ESTIMATE or SCENARIO]`","Tagged inputs, formula, range, sensitivity, overlap, and limitations","Only as a labeled scenario or break-even threshold anchored to an eligible factual/calculated hook"],["Inference `[I]`","Supported claims, mechanism, strongest alternative, and disconfirming check","Only adjacent to an eligible hook, explicitly bounded, and never as the subject-line fact"],["Hypothesis `[H]`","Testable, falsifiable, metric/guardrail/stop rule","Only as a permission-based diagnostic question anchored to an eligible hook"],["Complaint/allegation","E2 bounded experience/allegation or a reviewed aggregate calculation; never adjudicated fact","Individual allegations are excluded; aggregate themes require the public-corpus rule and must not be weaponized"],["Security signal","Narrow passive observation with technical sensitivity controls","Excluded from sales outreach; use a separate responsible-disclosure route when warranted"],["Unknown `[U]` or contradicted `[X]`","State explicitly; preserve reason/evidence","Never a factual hook"]],"line":215,"id":"03.13-evidence-use-matrix.t1"},{"type":"paragraph","text":"E0/E1 are hard-blocked from publication and activation. No matrix row overrides law, source rights, privacy, security, defamation, suppression, or human-review gates.","line":227}]}