From a4d29b2e71cc493a3a46c23226e376c375cf7c41 Mon Sep 17 00:00:00 2001 From: Leonard Date: Fri, 7 Aug 2026 05:17:09 +0000 Subject: [PATCH] =?UTF-8?q?docs:=20reconcile=20framework=20doc=20addendum?= =?UTF-8?q?=20+=20=C2=A79=20and=20Artifact=20A=20CVI=20note=20to=20two-cat?= =?UTF-8?q?egory=20GLOSSARY=20rule?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit --- docs/agents/audit-findings-template.md | 2 +- .../DOP-AI-Visibility-Unified-Framework-v1.md | 19 ++++++++++++------- 2 files changed, 13 insertions(+), 8 deletions(-) diff --git a/docs/agents/audit-findings-template.md b/docs/agents/audit-findings-template.md index a0f46da..daafa91 100644 --- a/docs/agents/audit-findings-template.md +++ b/docs/agents/audit-findings-template.md @@ -64,7 +64,7 @@ Describe the sampling method, engine set, and limitations. (Not gated — this i ## 3. Core Visibility Index (quantitative, 70 pts) -Directly countable from captured audit runs only. No qualitative bands folded in. +Directly countable from captured audit runs only. No qualitative bands folded in. Note: "quantitative" describes measurement precision, not evidentiary tier — this entire index is composed of Findings (per GLOSSARY.md's Tier 1/Evidence Model reconciliation) and stays Tier 2/Indicative regardless of how exactly it's counted. | Component | Weight | Score | |---|---:|---:| diff --git a/docs/architecture/DOP-AI-Visibility-Unified-Framework-v1.md b/docs/architecture/DOP-AI-Visibility-Unified-Framework-v1.md index bb69210..e5886f3 100644 --- a/docs/architecture/DOP-AI-Visibility-Unified-Framework-v1.md +++ b/docs/architecture/DOP-AI-Visibility-Unified-Framework-v1.md @@ -33,12 +33,17 @@ Resolution: Delete the "Observed/Inferred" rename — it solved a problem that d One real distinction worth preserving, not as a rename but as a note: the playbook's Tier 2 default rule — "Public observation = Tier 2 until owner access or stronger verification upgrades it" — is a *procedural* rule about how tiers get assigned, not a different taxonomy. Module A's Evidence Record schema (Section 20) should carry this default explicitly: any fact sourced from an AI engine output or public scrape starts at Tier 2 regardless of how confidently it was captured, and only moves to Tier 1 if the client confirms it directly. -Addendum — evidence authority nuance (added after Qwen review). The single Tier 1/2 scale holds, but different fact classes have different verifying authorities, and conflating them causes real confusion — "we verified the audit run happened" is not the same claim as "we verified the business fact inside it is true." Three classes: - -1. Business-entity facts (name, phone, address, hours, services, categories, profile ownership). Default Tier 2 when observed publicly. Promotes to Tier 1 only via owner confirmation or owner-access verification. -2. Audit-run facts (prompt executed, engine captured, timestamp recorded, raw output stored, competitor appeared in captured output). Default Tier 1 when supported by stored filesystem/git/database artifacts — this verifies the audit *event* occurred, not that the business fact inside the output is true. -3. Public claims inside AI outputs (an engine stating the client's phone number, hours, or service scope). Default Tier 2 until owner confirmation or owner-access verification upgrades it. -Action: add this three-class breakdown to GLOSSARY under the Tier 1/2 entry, so "verified" is never ambiguous between "the run happened" and "the fact is true." +Addendum — SUPERSEDED (2026-08-06). The three-class breakdown below was +wrong: it listed "competitor appeared in captured output" as a Tier +1-eligible "audit-run fact," but that's a claim about AI-output content, +not about whether the test ran. Reconciled version now lives in +GLOSSARY.md under "## Tier 1 — Verified" — two categories (Process facts +vs. Findings), not three classes. Findings, including anything about +mention status or competitor appearance, are Indicative by definition +per the pre-existing Evidence Model section and stay that way regardless +of how the audit event itself is verified. See GLOSSARY.md for the +authoritative version; this addendum is kept here only as a record of +the error and its correction, not as current guidance. --- @@ -298,7 +303,7 @@ Net effect: the critique doesn't change what gets built next, but it changes wha ## 9. What's now resolved vs. still open Resolved against the actual repo — audit-playbook-v1.md, README, path-to-poc-sequencing.md, and the four existing client/template files, all pulled from main (2026-08-06): -- No evidence-model collision existed. Verified = Tier 1, Indicative = Tier 2, one vocabulary, now with a three-class authority breakdown (business-entity facts / audit-run facts / public claims inside AI outputs). +- No evidence-model collision existed. Verified = Tier 1, Indicative = Tier 2, one vocabulary, now with a two-category reconciliation (process facts vs. Findings) — see GLOSSARY.md. - Layers 1–7 are the full, verified Path to PoC engagement sequence. The Stage↔Layer bridge map is confirmed, not a working hypothesis — including the Stage 4 double-duty correction (Layer 1b + Layer 4 both land there) that the sequence doc's own ordering forces. - Layer 7/POC Delivery confirmed at Stage 5, not Stage 6 — the sequence doc's own naming confirms this was correct, not just a reasonable guess. - Artifact A and Artifact B are trimmed to hold only content with no existing repo home, after discovering they originally duplicated data-inventory-v1.md, threat-register-v1.md, requirements-v1.md, and workflows-v1.md. Both now carry explicit Feed-Forward sections naming exact file paths, verified against live client directories (`docs/clients/overcome-fitness/`, `docs/clients/phoenix-salons/`).