Compatibility Commitments¶
Compatibility in bijux-canon-reason should be explicit: stable commands, tracked schemas,
durable artifacts, and release notes that explain intentional breakage.
This page should leave readers with a realistic sense of the compatibility bar. It is more valuable to be clear about what triggers review than to sound generously stable while leaving the real boundary ambiguous.
Treat the interfaces pages for bijux-canon-reason as the bridge between implementation detail and caller expectation. They should show what the package is prepared to defend before a dependency forms.
Visual Summary¶
flowchart LR
page["Compatibility Commitments<br/>clarifies: identify contracts | see caller impact | review compatibility"]
classDef page fill:#dbeafe,stroke:#1d4ed8,color:#1e3a8a,stroke-width:2px;
classDef positive fill:#dcfce7,stroke:#16a34a,color:#14532d;
classDef caution fill:#fee2e2,stroke:#dc2626,color:#7f1d1d;
classDef anchor fill:#ede9fe,stroke:#7c3aed,color:#4c1d95;
classDef action fill:#fef3c7,stroke:#d97706,color:#7c2d12;
surface1["HTTP app in src/bijux_canon_reason/api/v1"]
surface1 --> page
surface2["schema files in apis/bijux-canon-reason/v1"]
surface2 --> page
surface3["CLI app in src/bijux_canon_reason/interfaces/cli"]
surface3 --> page
proof1["reasoning traces and replay diffs"]
page --> proof1
proof2["apis/bijux-canon-reason/v1/schema.yaml"]
page --> proof2
proof3["apis/bijux-canon-reason/v1/pinned_openapi.json"]
page --> proof3
review1["tests/e2e for API, CLI, replay gates, retrieval reasoning, and smoke coverage"]
review1 -.raises compatibility pressure on.-> page
review2["tests/perf for retrieval benchmark coverage"]
review2 -.raises compatibility pressure on.-> page
review3["tests/unit for planning, reasoning, execution, verification, and interfaces"]
review3 -.raises compatibility pressure on.-> page
class page page;
class surface1,surface2,surface3 positive;
class proof1,proof2,proof3 anchor;
class review1,review2,review3 caution;
Compatibility Anchors¶
- README.md
- CHANGELOG.md
- pyproject.toml
Review Rule¶
Breaking changes must be visible in code, docs, and validation together.
Concrete Anchors¶
- CLI app in src/bijux_canon_reason/interfaces/cli
- HTTP app in src/bijux_canon_reason/api/v1
- schema files in apis/bijux-canon-reason/v1
- apis/bijux-canon-reason/v1/schema.yaml
Use This Page When¶
- you need the public command, API, import, schema, or artifact surface
- you are checking whether a caller can safely rely on a given entrypoint or shape
- you want the contract-facing side of the package before building on it
Decision Rule¶
Use Compatibility Commitments to decide whether a caller-facing surface is explicit enough to depend on. If the surface cannot be tied back to concrete code, schemas, artifacts, examples, and tests, treat it as unstable until that evidence is visible.
What This Page Answers¶
- which public or operator-facing surfaces
bijux-canon-reasonis really asking readers to trust - which schemas, artifacts, imports, or commands behave like contracts
- what compatibility pressure a change to this surface would create
Reviewer Lens¶
- compare commands, schemas, imports, and artifacts against the documented surface one by one
- check whether a seemingly local change actually needs compatibility review
- confirm that examples still point to real entrypoints and not to stale habits
Honesty Boundary¶
This page can identify the intended public surfaces of bijux-canon-reason, but real compatibility depends on code, schemas, artifacts, examples, and tests staying aligned. If those disagree, the prose is wrong or incomplete.
Next Checks¶
- move to operations when the caller-facing question becomes procedural or environmental
- move to quality when compatibility or evidence of protection becomes the real issue
- move back to architecture when a public-surface question reveals a deeper structural drift
Purpose¶
This page describes what should trigger compatibility review for the package.
Stability¶
Keep it aligned with the package's actual public surfaces and release process.