Audit: QRS - Goldenseal for Health & Longevity

Audit conducted on 11/17/2026 13:59 using AI4L + evidence-mcp / Opus 5.5

Iterations

Summary

Items Count
Total 97
Passed 89
Failed 0
N/A 8
Pass Rate 100.00%
  • Total = Passed + Failed + N/A
  • Pass Rate = Passed / (Passed + Failed) × 100
  • N/A items are excluded from the pass rate calculation

1. General Rules

# Description Result Comments
1.1 Every claim, magnitude, label, recommendation, and statement in the QRS is literally supported by content in the source ER. 🟢 Lede, Protocol, Time to effect, gates, tiers, markers and cadence all trace to ER text
1.2 Where the ER uses cautious phrasing (“not formally studied”, “None documented in human trials to date”, “theoretical concern”, “data are limited”), the QRS uses the same phrasing. 🟢 “for short periods without serious harmful effects”, “without a control group”, “without monitoring” kept as in ER
1.3 The QRS never strengthens an ER claim (e.g., “not formally studied” → “not required”) or softens one (e.g., “do not use during pregnancy” → “use with caution during pregnancy”). 🟢 No strengthening or softening found
1.4 The QRS does not relabel an ER fact under a different decision category. A “Benefit-Modifying Factor” from ER section is not surfaced as a “Caution”; a “Risk-Modifying Factor” is not surfaced as a “Side Effect”; etc. 🟢 Items sit under their ER categories
1.5 PubMed IDs, study citations, expert names, clinical trial identifiers (NCT*), and brand names appear in the QRS only if they appear in the source ER for the same fact. 🟢 No citations, PMIDs, NCT IDs or brands on the sheet

2. Focus, Tone & Audience

# Description Result Comments
2.1 The QRS follows the tone of the ER, which is determined by the ER’s own language, phrasing, and framing. 🟢  
2.2 The tone of the QRS is expert, accessible, objective, and data-driven 🟢  
2.3 The QRS reads as a trusted, knowledgeable guide rather than a prescriptive doctor 🟢  
2.4 The QRS avoids language that implies medical or clinical advice 🟢  
2.5 The QRS “presents information” instead of “providing guidance”, “recommending”, or “advising” 🟢  
2.6 The QRS never addresses the reader directly — it presents evidence, not guidance 🟢 No direct address
2.7 The QRS is written in plain language, avoiding unnecessary medical jargon 🟢 Technical terms (CYP3A4 substrates, hypernatremia, narrow-margin) are ER labels carried verbatim
2.8 Information is presented in a concise and very compact manner 🟢  
2.9 The target audience is health- and longevity-oriented adults who are risk-aware, proactive, and actively seeking to optimize health or apply the intervention under review. 🟢  
2.10 The target audience is willing to employ lifestyle and behavioral changes as well as follow protocols that may be inconvenient, costly, or require effort. 🟢  
2.11 The document is NOT written for the general population, who are unwilling to employ lifestyle and behavioral changes or follow protocols that may be inconvenient, costly, or require effort. 🟢  
2.12 Framing, takeaways, and risk/benefit weighting throughout the document reflect this audience, including where an intervention’s signal for the average person differs from its signal for this audience. 🟢 Lede weighs interaction risk for people on regular medications
2.13 The document’s own voice frames usage in longevity terms, not “anti-aging” (e.g., “anti-aging clinics”, “anti-aging community”, “anti-aging medicine”). Proper names that contain “anti-aging” (e.g., “American Academy of Anti-Aging Medicine”) are quoted verbatim. 🟢 No “anti-aging”
2.14 The document’s own voice uses formal clinical and scientific terminology, not colloquial or consumer-grade language (e.g., “oral medication” not “pill(s)”; “oral” / “administered orally” not “taken by mouth” / “given by mouth”; “injection” not “shot”; “adverse event” not “bad reaction”). This holds on EVERY surface, including the QRS lede — the stricter plain-language bar (see 2.7) does not license lay phrasing for route of administration. Direct quotes from sources are exempt. 🟢 No lay route or consumer phrasing

3. Template Integrity

# Description Result Comments
3.1 The following labels and headings on the QRS are fixed and not modified: Card and section headings: “Protocol”, “Time to effect”, “Benefits”, “Risk & Side Effects”, “Monitoring”, “Qualitative Assessment”; Gate headings: “Contraindications”, “Key Interactions”; Tier labels: “High”, “Medium”, “Low”, “Speculative”; Table column headers in Monitoring: “Marker”, “Target”, “Why”; 🟢 All fixed labels unchanged, incl. “Risk & Side Effects” (line 588)
3.2 Every element carrying a data-qrs-var attribute in [qrs_template] is present in the QRS with the same name 🟢 All template variables present; marker_#/qualitative_item_# expanded per 14.2/15.2
3.3 Everything outside the data-qrs-var spans — markup, text, styles, comments — is as in the template; spans that no checklist item addresses are left unchanged as well 🟢 Diff against template shows changes only inside variable spans and metadata values

4. Formatting

# Description Result Comments
4.1 An empty ER section never produces visible empty-state text in the QRS: the corresponding region is hidden or left empty. Phrases such as “None documented in human trials to date” or “Not formally studied” appear only where the ER uses them for a fact the QRS carries over. 🟢 Empty High/Medium benefit tiers and time_3 hidden, no empty-state text
4.2 Where the ER presents a bulleted item as “Label: content”, the QRS uses the ER’s bold label verbatim as the cell or row label. 🟢 Protocol labels, gate labels and marker names use ER bold labels
4.3 Labels are not paraphrased, abbreviated, or invented. 🟢 Only example lists trimmed (e.g., tacrolimus, tamoxifen dropped from parentheticals)
4.4 The QRS DOES NOT use emoji indicators (no 🟩, 🟥, 🟨, etc.). Color and emphasis are conveyed through CSS and bold labels. 🟢 No emoji
4.5 The QRS is designed to render on one A4 page. It fails this item only for one of four things: a trailing clause after a dash, an optional gloss, an example list left untrimmed, or a repetition. Nothing else is cut to fit. 🟢 No trailing dash clauses, optional glosses, untrimmed lists or repetitions found

5. Metadata

# Description Result Comments
5.1 The metadata is placed inside a single HTML comment that is the first element after “<!doctype html>” and before any other comment, head, or body content. 🟢  
5.2 Inside that HTML comment the YAML block is delimited by a line “—” opening and a line “—” closing. Text before the opening “—” is permitted but is not parsed as YAML. 🟢  
5.3 The metadata is not visible in any rendered view of the QRS and is not surfaced by any other element on the sheet. 🟢  
5.4 All frontmatter values are trimmed: no leading or trailing whitespace, no surrounding quotes unless the value contains a colon, bracket, or leading special character that requires YAML quoting. 🟢 duration quoted because it contains a colon
5.5 The filename of the source ER is stated as “er_filename: [er_filename]” 🟢  
5.6 Version of the QRS.md file used to create the document is stated as “qrs_prompt_version: [Version of QRS.md]” 🟢 26.11.9 matches QRS.md
5.7 Creation date and time of the document is stated as “qrs_creation_date: [YYYY-MMDD-HHMM]” (e.g., 2026-0501-1430) 🟢 2026-1117-1331
5.8 The nickname of the AI used to create the document is stated as “qrs_creator_ai_nickname: [qrs_creator_ai_nickname]” 🟢  
5.9 The nickname of the AI is the full name without the version number, with all remaining words joined by hyphens (e.g., Opus, Sonnet, Grok, Gemini-Flash, GPT); when the full name has no version number or is “Unknown”, the nickname equals the full name 🟢 Opus
5.10 The full name of the AI used to create the document is stated as “qrs_creator_ai_fullname: [qrs_creator_ai_fullname]” 🟢  
5.11 The full name of the AI consists of the model name (the word or words directly before the version number, leaving out the first if there is more than one, e.g., Claude Opus 5.5 → Opus 5.5), the model version number and the variant (if any) as stated in the system prompt or developer instructions of the creating model, with every word capitalized and without product, effort or pricing names (e.g., Opus X.Y, Sonnet X.Y, Grok X.Y, Gemini X.Y Flash, GPT X.Y), the model name alone when the system prompt or developer instructions name a model without a version number (e.g., Grok), or “Unknown” when neither the system prompt nor the developer instructions of the creating model name a model 🟢 Opus 5.5
5.12 The filename of the document is stated as “qrs_filename: [filename of this document]” 🟢  

6. Page Title & Header

# Description Result Comments
6.1 [page_title] is set to the [canonical_topic] of the ER frontmatter followed by “ - Quick Reference Sheet” (e.g., “Intervention - Quick Reference Sheet”). The [canonical_topic] is HTML-entity-encoded as needed (e.g., &amp; for &) 🟢  
6.2 [header_topic] is set to the [canonical_topic] of the ER frontmatter, with HTML entities encoded as needed (e.g., &amp; for &) 🟢  
6.3 [header_subline_date] is set to [qrs_creation_date reformatted as MM/DD/YYYY] 🟢 11/17/2026
6.4 [header_subline_model] is set to [qrs_creator_ai_fullname] 🟢 Opus 5.5
6.5 No additional header content appears: no badge, version stamp, AKA / alternate names line, source-AI attribution, audit date, or QRS variant marker. 🟢  

7. At-A-Glance Section

# Description Result Comments
7.1 [at_a_glance] opens by saying what the intervention is — its kind (e.g., dietary supplement, prescription medication, peptide, plant, procedure, practice) — and what it is used for, in plain language, before any verdict on the evidence 🟢 Opens with “a North American root herb sold as a dietary supplement and traditionally used for infections and digestive complaints”
7.2 [at_a_glance] is dense, execution-oriented summary of the ER Conclusion section 🟢 Condenses the Conclusion: berberine-borrowed benefits, interaction risk dominates for medication users
7.3 [at_a_glance] is no longer than 70 words 🟢 69 words
7.4 Every fact in [at_a_glance] is supported by a distinct passage in the ER. 🟢 Each fact maps to ER Motivation/History/Practical Considerations/Conclusion passages
7.5 It DOES NOT use acronyms or technical classifications that require specialist knowledge, uses plain-language terms instead 🟢  
7.6 It DOES NOT cite specific trials (names, years, sample sizes, p-values) 🟢  
7.7 It DOES NOT cite effect sizes, relative risks, or statistical results 🟢  

8. Contraindications

# Description Result Comments
8.1 The section is derived from the ER Key Interactions & Contraindications section 🟢  
8.2 [stop_items] represent the Contraindications from the ER 🟢 All four ER “Populations who should avoid” items present
8.3 Individual [stop_items] are formatted as <li></li> 🟢  
8.4 Items are as concise as possible. No trailing explanations, no elaborations, no mechanistic rationale, no attributions, no citations, no study details. No content after an em-dash, en-dash, or hyphen-dash (e.g., “— dose reduction required”, “— reduced efficacy”) — these trailing clauses are stripped. Just the key fact. 🟢  
8.5 Parenthetical qualifiers from the ER bullet — time windows, severity classes, threshold values, clinical staging — ARE preserved as part of the item, kept as concise as possible (shortened or trimmed where needed to fit the one-page budget, but never dropped entirely). 🟢 “on cyclosporine or tacrolimus”, “without monitoring” and an example list kept
8.6 When the ER uses ranking notation inside parens (e.g., “>” for severity ordering) that depends on an explanatory phrase to interpret, normalize the items to a plain comma-separated list rather than carrying through the bare symbol. 🟢 No ranking notation in ER
8.7 The section is left empty ONLY IF the ER’s Key Interactions & Contraindications section identifies no population, condition, or scenario for which the intervention should be avoided, deferred, or used only under specialist supervision. 🟢 Section populated; ER names populations to avoid
8.8 If the section is left empty, an HTML comment inside the [stop_items] span records the ER basis for the absence, naming or quoting the ER text relied on. E.g. N/A Section not empty

9. Key Interactions

# Description Result Comments
9.1 The section is derived from the ER Key Interactions & Contraindications section 🟢  
9.2 [caution_items] represent the ER’s documented interactions — the items under its “Documented interactions:” label — excluding any that are already listed as Contraindications and any entry phrased as no interaction found 🟢 Six documented interactions; four “no interaction found” entries excluded
9.3 [caution_theoretical_items] represent the ER’s theoretical interactions — the items under its “Theoretical interactions (inferred from mechanism, not studied in humans):” label 🟢 Four theoretical items
9.4 The ER placeholder “None identified” is never an item: it means that group is empty 🟢  
9.5 Individual [caution_items] and [caution_theoretical_items] are formatted as <li></li>. The template’s bold “Theoretical (inferred from mechanism)” sub-heading is the only marker of the second list’s status: no item carries “(theoretical)”, “inferred from” or a mechanism clause 🟢  
9.6 Items are as concise as possible. No trailing explanations, no elaborations, no mechanistic rationale, no attributions, no citations, no study details. No content after an em-dash, en-dash, or hyphen-dash (e.g., “— dose reduction required”, “— reduced efficacy”) — these trailing clauses are stripped. Just the key fact. 🟢  
9.7 Parenthetical qualifiers from the ER bullet — example drug lists, time windows, severity classes, threshold values, clinical staging — ARE preserved as part of the item, kept as concise as possible (shortened or trimmed where needed to fit the one-page budget, but never dropped entirely). 🟢 Example drug lists kept, trimmed
9.8 When the ER uses ranking notation inside parens (e.g., “>” for severity ordering) that depends on an explanatory phrase to interpret, normalize the items to a plain comma-separated list rather than carrying through the bare symbol. 🟢 No ranking notation in ER
9.9 If the ER’s theoretical group is empty, the whole Theoretical block — the caution_theoretical span with its sub-heading and list — is hidden with style="display:none", not removed: its spans stay in place, empty. N/A ER theoretical group is not empty
9.10 The section is left empty ONLY IF the ER’s Key Interactions & Contraindications section identifies no interaction, additive effect, or exposure that changes how the intervention is used. 🟢 Section populated
9.11 If the section is left empty, an HTML comment inside the [caution_items] span records the ER basis for the absence, naming or quoting the ER text relied on, and the Theoretical block is hidden per 9.9. E.g. N/A Section not empty

10. Protocol

# Description Result Comments
10.1 The section is derived from the ER Therapeutic Protocol section 🟢  
10.2 The three sets of [action] items cover the three most important actionable implementation aspects from that section 🟢 Trial regimen, short-term upper exposure, best time of day
10.3 If fewer than three distinct actionable implementation aspects are mentioned in the ER, the unused sets are left empty and hidden with style="display:none", not filled with placeholder text or empty-state phrasing. N/A ER has at least three actionable aspects
10.4 All used [action_#label], [action#value], [action#_sub] items are filled with meaningful content derived from that section. 🟢  

11. Time to Effect

# Description Result Comments
11.1 The three sets of [time] items cover the three most important time-to-effect aspects from the ER 🟢 ER Time to effect names two aspects (14-day enzyme inhibition, 28-day HbA1c change); both used
11.2 The sets are picked and ordered by the magnitude of the related benefit 🟢 Benefit-related HbA1c change first
11.3 If fewer than three distinct time-to-effect aspects are mentioned in the ER, the unused sets are left empty and hidden with style="display:none", not filled with placeholder text or empty-state phrasing. 🟢 time_3 spans hidden with style=”display:none”
11.4 All used [time_#label], [time#value], [time#_sub] items are filled with meaningful content derived from the ER. 🟢  
11.5 If the ER provides no information on time to effect at all, the whole Time to effect block (its subhead and all three sets) is hidden with style="display:none", not removed: its spans stay in place, empty. N/A ER provides time-to-effect information

12. Benefits

# Description Result Comments
12.1 The section is derived from the ER Expected Benefits section 🟢  
12.2 Key variables are [benefits_high], [benefits_medium], [benefits_low], [benefits_speculative] 🟢  
12.3 Items are as concise as possible. No explanations, no elaborations, no effect sizes, no qualifiers, no attributions, no citations, no study details, no mechanistic explanations, etc. Just the key fact. 🟢  
12.4 Parenthetical content — including effect sizes, sample notes, mechanistic hints, and example studies — is stripped, NOT preserved. 🟢 ER flags and parentheticals stripped
12.5 If a tier (high, medium, low, speculative) has no items, its span is hidden with style="display:none", not filled with “None documented in human trials to date” or similar empty-state phrasing. 🟢 High and Medium spans hidden

13. Risks

# Description Result Comments
13.1 The section is derived from the ER Potential Risks & Side Effects section 🟢  
13.2 Key variables are [risks_high], [risks_medium], [risks_low], [risks_speculative] 🟢  
13.3 Items are as concise as possible. No explanations, no elaborations, no effect sizes, no qualifiers, no attributions, no citations, no study details, no mechanistic explanations, etc. Just the key fact. 🟢  
13.4 Parenthetical content — including frequencies, severity grades, sample notes, mechanistic hints, and example studies — is stripped, NOT preserved. 🟢 “(Light-Triggered)”, “(High Blood Sodium)”, “(Kernicterus)” stripped
13.5 If a tier (high, medium, low, speculative) has no items, its span is hidden with style="display:none", not filled with “None documented in human trials to date” or similar empty-state phrasing. N/A All risk tiers have items

14. Monitoring

# Description Result Comments
14.1 The section is derived from the ER Monitoring Protocol & Defining Success section 🟢  
14.2 All measurable/quantifiable biomarkers from that section are listed, one table row per marker; the template’s # row is the pattern, repeated with # replaced by 1, 2, 3 … in the data-qrs-var names 🟢 All four ER table markers listed
14.3 [marker_#name] is the marker, [marker#target] its target value or range as given in the ER, [marker#_why] its purpose in a few words; a target the ER does not give is written “Not specified in the ER” 🟢 Targets as in ER
14.4 [monitoring_cadence] is populated with the monitoring cadence/frequency derived from that section. It is not left with placeholder text or empty. 🟢  
14.5 If the ER names no measurable biomarker, the table is hidden with style="display:none" and the single # row is left as in the template N/A ER names measurable biomarkers

15. Qualitative Assessment

# Description Result Comments
15.1 The section is derived from the ER Monitoring Protocol & Defining Success section 🟢  
15.2 All subjective/qualitative markers from that section are listed, one <li> per marker; the template’s # item is the pattern, repeated with # replaced by 1, 2, 3 … 🟢 All five qualitative markers listed
15.3 If the ER names no qualitative marker, the card’s list is hidden with style="display:none" and the single # item is left as in the template N/A ER names qualitative markers

Issues 11/17/2026 13:59

Pass rate 100.00%. No issues found.

Issues 11/17/2026 13:55

  1. 4.5 — Untrimmed example lists kept: Three parenthetical example lists carry every ER example: Contraindications line 551 “(flecainide, carbamazepine, tamoxifen)”, Theoretical line 574 “(insulin, glipizide, glyburide)” and line 576 “(cinnamon, chromium, alpha-lipoic acid)”.

Fixes 11/17/2026 13:55

  1. 4.5 — Example lists trimmed: Trimmed Contraindications “(flecainide, carbamazepine, tamoxifen)” to “(flecainide, carbamazepine)”, and Theoretical “(insulin, glipizide, glyburide)” to “(insulin, glipizide)” and “(cinnamon, chromium, alpha-lipoic acid)” to “(cinnamon, chromium)”.

Issues 11/17/2026 13:51

  1. 2.7 — Jargon in Protocol sub-line: The action_3_sub at line 478 ends with “hydrastine half-life about 4.8 hours”, an unexplained alkaloid name and pharmacokinetic term that is not part of any ER label carried verbatim.

Fixes 11/17/2026 13:51

  1. 2.7 — Jargon in Protocol sub-line: Removed the trailing “hydrastine half-life about 4.8 hours” from action_3_sub, leaving “Dose spacing used in a drug-interaction trial; split dosing two to three times daily is standard”.

Issues 11/17/2026 13:48

  1. 1.2 / 14.3 — Trough target hedge dropped: marker_4_target at line 655 reads “Transplant team’s individual target range”, omitting the ER’s “No established target; track against the transplant team’s individual target range” (ER Monitoring table), so the target is not as given in the ER and its cautious phrasing is lost.

Fixes 11/17/2026 13:48

  1. 1.2 / 14.3 — Trough target hedge restored: Changed marker_4_target from “Transplant team’s individual target range” to “No established target; transplant team’s individual target range”, matching the ER’s Monitoring table wording.

Issues 11/17/2026 13:44

  1. 2.7 — Specialist jargon in Protocol: action_3_sub at line 478 reads “Dose spacing used in the probe-cocktail trial”; “probe-cocktail” is unnecessary specialist jargon on a non-label surface.
  2. 4.5 — Repetition and untrimmed list: time_2_sub at line 506 (“Enzyme inhibition was present after 14 days of use”) only restates its label and value, and the theoretical item at line 574 carries the ER’s four-example list “(berberine, Oregon grape, barberry, goldthread)” untrimmed.
  3. 9.2 — Cyclosporine interaction missing: The ER’s documented “Cyclosporine” interaction (ER line 323) is absent from caution_items (lines 558-564), although the contraindication at line 550 covers only organ-transplant recipients.
  4. 10.4 — Sub drawn from other section: action_2_sub at line 467 appends “long-term safety is uncertain”, which comes from the ER Discontinuation & Cycling section, not the Therapeutic Protocol section.

Fixes 11/17/2026 13:44

  1. 2.7 — Specialist jargon in Protocol: Replaced “Dose spacing used in the probe-cocktail trial” with “Dose spacing used in a drug-interaction trial” in action_3_sub.
  2. 4.5 — Repetition and untrimmed list: Replaced the repetitive time_2_sub with “Shown as raised midazolam levels; activity returns only as new enzyme is made”, and trimmed the berberine-containing supplements example list to “(berberine, Oregon grape)”.
  3. 9.2 — Cyclosporine interaction missing: Added “Cyclosporine” to caution_items after “Metformin”, following the ER’s order.
  4. 10.4 — Sub drawn from other section: Removed “; long-term safety is uncertain” from action_2_sub, leaving “For short periods without serious harmful effects”.

Issues 11/17/2026 13:40

  1. 1.3 — Trial timing presented generally: Protocol cell 3 (lines 472–478) gives “About 8 am, 12 pm and 4 pm” as the best time of day without the ER’s qualifier that this was the dose spacing “in the probe-cocktail trial”.
  2. 1.4 — Recovery fact under Time: time_2_sub (line 506) “Activity returns only as new enzyme is made” is the ER’s Discontinuation & Cycling recovery fact; next to “14 days” it misstates the ER’s Time to effect fact that inhibition “was present after 14 days”.

Fixes 11/17/2026 13:40

  1. 1.3 — Trial timing qualifier restored: Prefixed action_3_sub with “Dose spacing used in the probe-cocktail trial;” so the 8 am / 12 pm / 4 pm schedule carries the ER’s trial qualifier.
  2. 1.4 — Time-to-effect sub corrected: Replaced time_2_sub “Activity returns only as new enzyme is made” (a discontinuation fact) with the ER’s time-to-effect wording “Enzyme inhibition was present after 14 days of use”.

Issues 11/17/2026 13:36

  1. 4.5 — Repeated drugs and restated sub: Tacrolimus (line 550 / 559) and tamoxifen (line 551 / 560) each appear in both a contraindication and a Key Interactions example list, and time_2_sub “Inhibition present after 14 days of use” (line 506) only restates its label and value.

Fixes 11/17/2026 13:36

  1. 4.5 — Repeated drugs and restated sub: Removed tacrolimus from the CYP3A4 example list and tamoxifen from the CYP2D6 example list (both already named in Contraindications), and replaced time_2_sub “Inhibition present after 14 days of use” with the ER’s “Activity returns only as new enzyme is made”.

Issues 11/17/2026 13:33

  1. 9.6 — Metformin item elaborated: The Key Interactions item at line 561 reads “Metformin (exposure lowered most at low metformin doses)”, adding an explanation of the interaction effect beyond the ER bold label “Metformin:”.
  2. 11.3 — Non-onset third time set: Time to effect set 3 (lines 511–517, “Enzyme recovery / 30-day washouts”) is a washout detail from Discontinuation & Cycling, not a time-to-effect aspect; the ER names only two, so set 3 should be empty and hidden.

Fixes 11/17/2026 13:33

  1. 9.6 — Metformin item elaborated: Reduced the Key Interactions item from “Metformin (exposure lowered most at low metformin doses)” to the ER bold label “Metformin”.
  2. 11.3 — Non-onset third time set: Emptied the time_3_label, time_3_value and time_3_sub spans (“Enzyme recovery / 30-day washouts / …”) and hid each with style=”display:none”.