Skip to content

chore: Fabric pin を v0.9.0rc4 へ上げる（AZ-06 は EffectObservation の生産者） #49

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
01rabbit merged 1 commit into from
Sep 20, 2026
Merged

chore: Fabric pin を v0.9.0rc4 へ上げる（AZ-06 は EffectObservation の生産者） #49

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
26 changes: 24 additions & 2 deletions docs/fabric-pin.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -30,7 +30,7 @@ drift structurally impossible rather than merely noticed.

AZ-06 pins the Azazel-Fabric release tag:

`v0.9.0rc3`
`v0.9.0rc4`

### Why a release candidate

Expand Down Expand Up @@ -59,6 +59,27 @@ ref, and `runtime/effect_projection.py` had to refuse them rather than flatten
them. Rewriting a producer's identifier is not an option: `surface:http-8080`
is a well-formed reference to a surface that does not exist.

`v0.9.0rc4` is the pin now. It narrows `EffectObservation` to the two
authority classes an observation can have observed — `observed_fact` and
`active_materialized` (Azazel-Fabric#52) — and refuses the four others that
`rc3` accepted. **It is the first non-additive candidate in the series**, and
the reason this repository moves to it is not that the grammar changed
somewhere: AZ-06 is a *producer* of `EffectObservation`, so this is the
contract it emits under.

The move costs nothing here because `runtime/effect_projection.py` already
derives the class from the observation's status and can emit only those two —
`observe_effect` has no parameter for it. What changes is that this became a
guarantee of the contract instead of a habit of one function. A future edit
introducing a third class would now be refused at the boundary rather than
accepted and carried downstream.

The pin was verified before the move rather than after: the `v0.9.0rc4` tag's
own checkout reproduces `release/v0.9.0rc4.digest.json`, whose signature
verifies under the release owner's key, and this repository's full suite runs
green against it. **What is signed is that digest; the git tag itself is
unsigned** and is not a trust boundary (Azazel-Fabric#56).

That refusal was recorded as a test designed to fail when the pin moved, and it
did. The stand-in values are gone from
`tests/test_effect_contracts_producer.py`; AZ-06's own references now go into
Expand Down Expand Up @@ -208,4 +229,5 @@ updates `pyproject.toml`, this document, and the CHANGELOG together.
| `c3154de` (2026-08-15) | `v0.5.0` → `v0.6.0` |
| `77ad0cd` (2026-08-22) | `v0.6.0` → `v0.8.0` |
| Deception#28 AC-7 (2026-09-20) | `v0.8.0` → `v0.9.0rc2` |
| `effect_contracts` producer (2026-09-20) | `v0.9.0rc2` → `v0.9.0rc3` (current) — the reference grammar; see above |
| `effect_contracts` producer (2026-09-20) | `v0.9.0rc2` → `v0.9.0rc3` — the reference grammar; see above |
| `EffectObservation` authority narrowing (2026-09-20) | `v0.9.0rc3` → `v0.9.0rc4` (current) — AZ-06 emits this record, so it moves with the contract it emits under; see above |
2 changes: 1 addition & 1 deletion pyproject.toml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -12,7 +12,7 @@ license = {text = "MIT"}
authors = [{name = "01rabbit"}]
dependencies = [
"PyYAML>=6.0",
"azazel-fabric @ git+https://github.com/01rabbit/Azazel-Fabric.git@v0.9.0rc3",
"azazel-fabric @ git+https://github.com/01rabbit/Azazel-Fabric.git@v0.9.0rc4",
]

[project.optional-dependencies]
Expand Down
Loading