docs: present AI strategy as a diagram-first story
Build and publish Eldorado documentation / build (pull_request) Successful in 40s

This commit is contained in:
jarvis-at-skic committed 2026-10-04 12:23:41 +00:00
1 parent 7f40de708c
commit 794f2aa2d6
30 files changed
+2496 -4

No files matched your search

+26
View File
@@ -2,11 +2,28 @@
id: skills-quality
title: Skills quality
description: Risk tiers, independent evaluation, repeated trials, adversarial conformance.
hide_table_of_contents: true
---
import ArchifyDiagram from '@site/src/components/ArchifyDiagram';
<div className="strategy-content">
**TO-BE · Target operating model — not installed controls or company results.**
## What evidence is enough to promote a skill?
Risk determines the evaluation depth, not whether safety checks apply. Independent evidence must show useful outcomes and no failed mandatory gate before promotion.
<ArchifyDiagram name="strategy-quality" title="Classify risk, test independently, then pass every mandatory gate" story />
- **R1:** approved nonsensitive drafting. **R2:** code or customer-impacting material. **R3:** privileged or sensitive effects; human-authorized, outside initial autonomous scope.
- **No averaging away harm:** a critical unauthorized effect blocks promotion.
- **Proposed floors need owner ratification:** missing coverage or acceptance rules means qualification is incomplete.
<details>
<summary>Evaluation design, adversarial cases and interpretation</summary>
**TO-BE · Target operating model.** Introduction depends on the [roadmap gates](./roadmap.mdx); these pages do not describe installed controls or measured company results.
Quality has four separate questions: was the right skill selected, was the procedure executed safely, did the artifact pass independent checks, and did the workflow produce useful outcomes? A successful invocation answers none of the remaining questions by itself.
@@ -21,6 +38,11 @@ Include adversarial repository text, malicious referenced code, wrong permission
Report denominators, task mix, trial counts, per-client results and uncertainty. A passing benchmark is evidence within a tested envelope, not proof that all future tasks are safe. Material model, client, adapter or dependency changes trigger requalification.
</details>
<details>
<summary>Proposed R1 / R2 / R3 sample sizes and thresholds</summary>
## Proposed qualification floors
These are local design defaults for responsible-owner ratification, not approved service levels or guarantees. All mandatory checks must pass regardless of aggregate score.
@@ -57,4 +79,8 @@ These are local design defaults for responsible-owner ratification, not approved
Artifact success denominator is every planned executed trial, retaining failures and abandonment. Trials of same task are correlated: show per-task and per-client strata; do not inflate independent sample size. Selection floors use independent labels. Report 95% uncertainty intervals clustered by task for outcome comparisons; no regression versus current approved path beyond a preregistered owner-approved margin. Missing margin/precision rule prevents promotion.
</details>
**Next:** [Introduce qualified skills through a bounded pilot →](./skills-adoption.mdx)
</div>