Skip to content

docs(decisions): agent-bridle publishing block — submodule analysis + 3 options#299

Merged
hartsock merged 1 commit into
mainfrom
docs/bridle-publishing-decision
Jun 12, 2026
Merged

docs(decisions): agent-bridle publishing block — submodule analysis + 3 options#299
hartsock merged 1 commit into
mainfrom
docs/bridle-publishing-decision

Conversation

@hartsock

Copy link
Copy Markdown
Member

What this PR does

Adds docs/decisions/agent_bridle_publishing.md answering the operator's question: does making brush a submodule inside agent-bridle sidestep the crates.io publishing block?

Short answer: no. A submodule relocates source files; it does not change the dependency edge crates.io rejects (path/git deps on unpublished crates are both forbidden at publish time). The doc explains that, then costs three real options:

Recommendation: C now, A if publishing pressure arrives first; B not recommended absent an external consumer needing published brush-backed agent-bridle before #1184. The usability half is already solved by #297 (--disable-ocap), so publishing is off the critical path.

Test plan

Docs-only — no code, gates trivially green. Decision awaits operator pick (status line in the doc).

Out of scope

🤖 Generated with Claude Code

… 3 options

Operator asked whether a brush submodule inside agent-bridle sidesteps
the crates.io publishing block. It does not: a submodule changes where
brush source lives, not the dependency edge crates.io rejects. Documents
that, then costs three real options (publish-clean-leaves-only /
vendor-brush-as-MIT-modules / wait-for-brush#1184) and recommends C-now /
A-if-pressured, with B (vendor/submodule) not recommended absent an
external consumer needing published brush-backed agent-bridle pre-#1184.

The usability half is already solved by #297 (--disable-ocap); this doc
is only about the publishing half.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
@hartsock
hartsock merged commit 8ab05c1 into main Jun 12, 2026
4 checks passed
@hartsock
hartsock deleted the docs/bridle-publishing-decision branch June 12, 2026 03:51
hartsock added a commit that referenced this pull request Jun 27, 2026
… 3 options (#299)

Operator asked whether a brush submodule inside agent-bridle sidesteps
the crates.io publishing block. It does not: a submodule changes where
brush source lives, not the dependency edge crates.io rejects. Documents
that, then costs three real options (publish-clean-leaves-only /
vendor-brush-as-MIT-modules / wait-for-brush#1184) and recommends C-now /
A-if-pressured, with B (vendor/submodule) not recommended absent an
external consumer needing published brush-backed agent-bridle pre-#1184.

The usability half is already solved by #297 (--disable-ocap); this doc
is only about the publishing half.

Co-authored-by: hartsock <hartsock@users.noreply.github.com>
Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant