0.3.0

Installer detection is now two-tier: agents installed on the system (Claude Code, OpenCode, OpenClaw) get their global files once; agents actually used in a repo get per-repo files only there. --all / --only adopt an agent in a fresh repo.

Full Changelog: v0.2.0...v0.3.0

0.2.0

What's Changed

• ci: fail PRs whose commits carry AI-authorship attribution by @pallaoro in #35
• fix(sidenote): park notes in .git/info, not a tracked file at the project root by @pallaoro in #36
• feat(guidelines): existence is not reachability — trace the current shape with the real input by @pallaoro in #37
• feat(skills): verify-responsive — render the UI at real viewport widths by @pallaoro in #38

Full Changelog: v0.1.0...v0.2.0

v0.1.0

What's Changed

• feat: add Skillify + DRY/MECE resolver skills (§6) and sharpen §2/§5 by @pallaoro in #1
• feat: add /scalable slash command by @pallaoro in #2
• feat: add principle 7 'Ground in Reality, Don't Recall' by @pallaoro in #3
• feat: add grounding step to /scalable command by @pallaoro in #4
• feat: cross-agent distribution machine + §1 ambiguity check by @pallaoro in #5
• docs: add The Greybeard mascot as README hero by @pallaoro in #6
• chore: rename to Greybeard by @pallaoro in #7
• docs: update Greybeard mascot art (darker render) by @pallaoro in #8
• feat: name the three pillars (Scalability / Long-term / Efficiency) + define-first step by @pallaoro in #9
• feat: company vision-pillars extend the three (never replace them) by @pallaoro in #10
• feat: reuse before you build — look down the stack by @pallaoro in #11
• feat: prefer docs-retrieval tools (Context7 / find-docs) for latest docs by @pallaoro in #12
• Sharpen /scalable grounding: per-claim, not per-session; do the cheap check yourself by @pallaoro in #13
• scalable: sweep the whole repo for the symptom before hypothesising where it lives by @pallaoro in #15
• scalable: reused state inherits every behavior attached to it by @pallaoro in #16
• guidelines: verify the invocation contract + "compiles ≠ works" (§7, §4) by @pallaoro in #17
• guidelines: give Greybeard a spine — recommend, disagree out loud (§1) by @pallaoro in #18
• feat(commands): add /sidenote to park a thought without derailing the current task by @pallaoro in #19
• fix(install): don't duplicate guidelines into a hand-merged CLAUDE.md by @pallaoro in #20
• feat(sidenote): self-contained capture + checkbox lifecycle; ignore the parking file by @pallaoro in #21
• rename /scalable command to /pressure-test by @pallaoro in #22
• ci: add npm publish workflow by @pallaoro in #23
• docs: document updating + releasing (main already added the publish workflow) by @pallaoro in #24
• feat(guidelines): §7 rule — the report is a symptom, not a diagnosis by @pallaoro in #28
• when getting a prompt from a by @pallaoro in #30
• refactor(pressure-test): main context is the main brain; subagents are scoped instruments by @pallaoro in #31
• feat: ruling-lifecycle rules — ledger, challenge protocol, behavior pricing by @pallaoro in #32
• feat(visualize): /visualize command — draw the real shape, anchored to file:line by @pallaoro in #33
• https github com dmmulroy dotfiles blob 7f5c3d03b4c432c1e by @pallaoro in #34

New Contributors

• @pallaoro made their first contribution in #1

Full Changelog: https://github.com/clawnify/greybeard/commits/v0.1.0