Context index
Why this project is built the way it is.
0
1
2
3
4
5
6
7
8
9
A
B
C
- compatibility.md — findings from testing Keep the Why alongside other skills/frameworks, and design decisions about the skill's own activation/setup-triggering behavior
- config-format.md — design of the skill's own setup/config mechanism (
.keep-the-why/~/.keep-the-why/files,context-schema, confirmation settings, pinned versions)
D
- dashboard.md — design of the read-only dashboard (
dashboard/): why it exists beside the "no dashboard" line, why in this repo, the plain-JS page, the fingerprint/SSE live mode, Git authors as-is
E
- entry-format.md — design of
context/entries and layout themselves: flat directory (no subdirectories), theTypefield, a found contradiction surfaced rather than superseded - evals.md — design of the eval runner and suite (
tools/evals/): package layout, what the judge and the mechanical checks each decide
F
G
H
I
- issue-triage.md — how reports are sorted: three issue forms,
skill-wordingvs.bug, theevalsorigin label
J
K
L
- lint.md — design decisions internal to the structural linter (
lint/): schema gating, severity model, parsing heuristics
M
N
O
P
- positioning.md — editorial/positioning choices for README,
llms.txt, and docs
Q
R
- release-and-distribution.md — how this repo builds, releases, and gets distributed/installed across ecosystems