concepts - grisuno/ReadMenator GitHub Wiki
Concepts
Second-brain semantic layer: nouns map atomically to file sets (EXTRACTED); verbs aggregate structural edges (INFERRED).
Verb Edges
| Source | Verb | Target | Strength |
|---|---|---|---|
graph |
depends_on |
readmenator |
1.00 |
file |
depends_on |
readmenator |
0.99 |
graph |
depends_on |
file |
0.93 |
readmenator |
depends_on |
file |
0.93 |
returns |
depends_on |
readmenator |
0.91 |
project |
depends_on |
readmenator |
0.87 |
build |
depends_on |
readmenator |
0.85 |
project |
depends_on |
file |
0.80 |
return |
depends_on |
readmenator |
0.80 |
returns |
depends_on |
file |
0.80 |
text |
depends_on |
readmenator |
0.80 |
map |
depends_on |
readmenator |
0.79 |
build |
depends_on |
file |
0.78 |
files |
depends_on |
readmenator |
0.78 |
graph |
depends_on |
every |
0.77 |
node |
depends_on |
readmenator |
0.77 |
edge |
depends_on |
readmenator |
0.76 |
file |
depends_on |
graph |
0.76 |
readmenator |
depends_on |
every |
0.76 |
readmenator |
depends_on |
source |
0.75 |
file |
depends_on |
every |
0.74 |
readmenator |
depends_on |
graph |
0.74 |
symbol |
depends_on |
readmenator |
0.73 |
edges |
depends_on |
readmenator |
0.72 |
file |
depends_on |
source |
0.72 |
files |
depends_on |
file |
0.72 |
graph |
depends_on |
source |
0.72 |
node |
depends_on |
file |
0.72 |
source |
depends_on |
readmenator |
0.72 |
all |
depends_on |
readmenator |
0.71 |
nodes |
depends_on |
readmenator |
0.70 |
readmenator |
depends_on |
all |
0.70 |
readmenator |
depends_on |
import |
0.70 |
returns |
depends_on |
graph |
0.70 |
empty |
depends_on |
readmenator |
0.69 |
symbol |
depends_on |
file |
0.69 |
file |
depends_on |
import |
0.68 |
per |
depends_on |
readmenator |
0.68 |
readmenator |
depends_on |
symbol |
0.68 |
text |
depends_on |
file |
0.68 |
config |
depends_on |
readmenator |
0.67 |
edges |
depends_on |
file |
0.67 |
graph |
depends_on |
import |
0.67 |
graph |
depends_on |
map |
0.67 |
returns |
depends_on |
every |
0.67 |
contract |
depends_on |
readmenator |
0.66 |
file |
depends_on |
map |
0.66 |
graph |
depends_on |
return |
0.66 |
project |
depends_on |
every |
0.66 |
when |
depends_on |
readmenator |
0.66 |
Dialectic Prompts
- Thesis:
allcentralizes 53 files; Antithesis:buildpulls 54 files with 28 shared (Jaccard 0.35); Synthesis: should they merge, split by layer, or keepbridgesexplicit? - Thesis:
allcentralizes 53 files; Antithesis:contentpulls 51 files with 24 shared (Jaccard 0.30); Synthesis: should they merge, split by layer, or keepbridgesexplicit? - Thesis:
allcentralizes 53 files; Antithesis:directorypulls 38 files with 22 shared (Jaccard 0.32); Synthesis: should they merge, split by layer, or keepbridgesexplicit? - Thesis:
allcentralizes 53 files; Antithesis:everypulls 53 files with 27 shared (Jaccard 0.34); Synthesis: should they merge, split by layer, or keepbridgesexplicit? - Thesis:
allcentralizes 53 files; Antithesis:filepulls 91 files with 43 shared (Jaccard 0.43); Synthesis: should they merge, split by layer, or keepdepends_onexplicit? - Thesis:
allcentralizes 53 files; Antithesis:filespulls 66 files with 34 shared (Jaccard 0.40); Synthesis: should they merge, split by layer, or keepbridgesexplicit? - Thesis:
allcentralizes 53 files; Antithesis:graphpulls 73 files with 33 shared (Jaccard 0.35); Synthesis: should they merge, split by layer, or keepdepends_onexplicit? - Thesis:
allcentralizes 53 files; Antithesis:listpulls 39 files with 27 shared (Jaccard 0.42); Synthesis: should they merge, split by layer, or keepbridgesexplicit? - Thesis:
allcentralizes 53 files; Antithesis:mappulls 47 files with 24 shared (Jaccard 0.32); Synthesis: should they merge, split by layer, or keepbridgesexplicit? - Thesis:
allcentralizes 53 files; Antithesis:nodespulls 55 files with 28 shared (Jaccard 0.35); Synthesis: should they merge, split by layer, or keepbridgesexplicit?