fiddlerwoaroof/improvise
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

chore: seed prolog knowledge base and track kb.pl

Browse Source 
AGENTS.md tells agents to load the `kb` Prolog session at session start,
but prolog-sessions/ was fully gitignored, so the KB could never survive a
fresh clone -- every session started from an empty knowledge base.

Narrow the ignore to prolog-sessions/* with an exception for kb.pl, and
check in a seed KB built from context/repo-map.md and
context/design-principles.md: crate/layer/coupling rules, the
"where do I change X" location table, the Cmd -> Effect pipeline,
the load-bearing invariants, testing doctrine, and the beads workflow.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01V69PGQfMJuGRmtSWpfVAqn
This commit is contained in:
Edward Langley
2026-08-29 10:25:01 -07:00
parent 605d412225
commit aabee3a460
2 changed files with 87 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files
.gitignore
+3 -1
View File
@@ -30,4 +30,6 @@ proptest-regressions/
/roadmap.html
# prolog-mcp session files (server writes ./prolog-sessions relative to cwd)
prolog-sessions/
# ...except the checked-in seed knowledge base, which agents load at session start.
prolog-sessions/*
!prolog-sessions/kb.pl