@javierfrancoleal/autonomy-grader 1.0.0 → 1.1.0
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/assets/autonomy-grader.md +10 -0
- package/package.json +1 -1
|
@@ -7,6 +7,16 @@ rules itself.
|
|
|
7
7
|
|
|
8
8
|
`Input → Detection (AI + human) → AI processes, decides, and proposes → Human reflects, verifies, and decides → Output`
|
|
9
9
|
|
|
10
|
+
## Evolution (variation + selection)
|
|
11
|
+
|
|
12
|
+
The grader is the fitness function; add the variation:
|
|
13
|
+
|
|
14
|
+
- When a tier has been stable for a while, propose 1-2 CANDIDATE rule
|
|
15
|
+
mutations (small changes to decisions.md) as YELLOW proposals.
|
|
16
|
+
- Let the next cycle's grades score the candidates; keep winners, discard
|
|
17
|
+
the rest — with the owner's approval, as always.
|
|
18
|
+
- Never mutate more than one rule at a time: evolution is incremental.
|
|
19
|
+
|
|
10
20
|
## Purpose
|
|
11
21
|
|
|
12
22
|
Over time, the supervision log accumulates evidence: which GREEN and
|
package/package.json
CHANGED
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
{
|
|
2
2
|
"name": "@javierfrancoleal/autonomy-grader",
|
|
3
|
-
"version": "1.
|
|
3
|
+
"version": "1.1.0",
|
|
4
4
|
"description": "Governance layer of the J.Franco agentic workflow: grades the reliability of automatic decisions from the supervision log and proposes tier moves — the human approves every change.",
|
|
5
5
|
"author": "J.Franco",
|
|
6
6
|
"license": "MIT",
|