7 Commits

Author SHA1 Message Date
d93820415f feat: symlink settings.json globally, update bootstrap and readme 2026-03-26 21:24:24 -05:00
02c2c2fabb docs: document sync hooks, gitattributes, global setup, and remotes 2026-03-26 20:49:25 -05:00
a0593e81a5 docs: add new machine setup instructions to README 2026-03-26 20:48:14 -05:00
Marcio Puga
65e5043c6f docs: clarify that the architecture evolves, not the model
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
2026-03-24 22:42:46 +11:00
Marcio Puga
654c0fdb75 docs: reframe Cog as convention-based learning tool with Unix-native design
Plain text by design — not as a compromise, but because it lets Claude
reason over its own memory with grep, find, and git diff. Positions Cog
as an experiment in watching how AI organizes itself given conventions
and self-observation capabilities.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
2026-03-24 22:37:01 +11:00
Marcio Puga
d47510b0f6 feat: scalable entities & patterns architecture
- Entity format: 3-line compact registry (### Name / key facts / status+links)
  Heavy entries promoted to thread files. Cross-domain pointers for shared entities.
- Pattern satellites: core patterns.md cap reduced from 110→70 lines (5.5KB).
  Domain-specific patterns go in satellite files loaded only by owning skill.
- Reflect: pattern routing rules, entity format enforcement (step 3b)
- Housekeeping: entity registry enforcement (step 5b), glacier inactive entities
- Evolve: scorecard metrics for pattern distribution + entity compression ratio
- README: updated entity example to 3-line format

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
2026-03-23 07:03:34 +11:00
Marcio Puga
1dd881975b feat: initial cog release — cognitive architecture for Claude Code
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
2026-03-18 20:04:29 +11:00