Release 1.2.0 — CLI peer-facts parity + enrichment/doctor fixes#4
Release 1.2.0 — CLI peer-facts parity + enrichment/doctor fixes#4tachyon-beep wants to merge 4 commits into
Conversation
… fixes Version bump + CHANGELOG ceremony for 1.2.0. The code (wardline-peer-facts and requirements-enrichment CLI subcommands, plus the rejected-trace enrichment-coverage fix and the root-aware Wardline doctor remediation) already landed on main; this branch carries only the version + changelog. Minor bump (not patch): the two new CLI subcommands are new public surface. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 02fdf2edb3
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
| @@ -1 +1 @@ | |||
| __version__ = "1.1.0" | |||
| __version__ = "1.2.0" | |||
There was a problem hiding this comment.
Update the deployed site for 1.2.0
When this release bump is tagged as v1.2.0, the package will report 1.2.0 but the public member site remains stale: I checked site/src/pages/index.astro, which still hard-codes VERSION = '1.0.0' and a hero note saying the snapshot is 1.1.0 untagged, and .github/workflows/site.yml only deploys on site/** changes. Because this commit changes only the package version/changelog, merging it leaves users seeing contradictory release information until a separate site change is made.
Useful? React with 👍 / 👎.
Brings release/1.2.0 (a stale ceremony cut from 8258f76, version-bump only) up to current main and folds in the work stranded off every release branch: - PDR-016 operator web UX + a11y overhaul (was on main, not here) - PDR-018 seam hardening (Loomweave/Legis/Filigree contract tests) - wheel-build force-include fix + PDR-017 producer-gated-live + consumer handoff Keeps the 1.2.0 version bump. No history rewrite. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 8c6a97e765
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
|
|
||
| from typing import Any | ||
|
|
||
| LOOMWEAVE_ADAPTER_STATUSES = {"available", "unavailable"} |
There was a problem hiding this comment.
Include degraded in Loomweave catalog statuses
When the Loomweave DB is present but partial, e.g. the schema exists without SEI tables or the catalog schema is incomplete, the producer intentionally returns adapter_status.status == "degraded" (src/plainweave/loomweave_adapter.py:536-543, and existing adapter tests assert that state). This new contract validator rejects every status except available/unavailable, so routing live catalog output through it will fail for a supported degraded producer state rather than pinning its shape/no-silent-clean behavior. Add degraded here and validate its expected in-band degradation fields.
Useful? React with 👍 / 👎.
| "intent/intent-coverage.json", | ||
| "loomweave/identity-resolve.json", | ||
| "loomweave/identity-sei.json", | ||
| "loomweave/catalog-degraded.json", |
There was a problem hiding this comment.
Expose the Loomweave catalog contract resource
This adds weft.plainweave.loomweave_catalog.v1 as a committed contract fixture, but the MCP contract catalog was not extended: plainweave_project_context_get(include_contracts=True) only advertises MCP_RESOURCE_URIS, and read_resource rejects anything not in CONTRACT_RESOURCES; both still omit plainweave://contracts/weft.plainweave.loomweave_catalog.v1. Consumers can call the catalog tool but cannot discover or read the newly pinned contract via MCP, so add the URI/resource and resource-inventory entry alongside this fixture.
Useful? React with 👍 / 👎.
…zed CHANGELOG, README) # Conflicts: # CHANGELOG.md
Release ceremony for 1.2.0. Diff is the two ceremony files only (
_version.py+CHANGELOG.md); all code already landed onmain.Minor bump (not patch): the two new CLI subcommands are new public surface.
Included since v1.1.0
Added
plainweave wardline-peer-facts [--json] [--limit N] [--offset N]— CLI parity for theweft.plainweave.wardline_peer_facts.v1envelope (previously MCP-only).plainweave requirements-enrichment <entity_ref>... [--json]— CLI parity for theweft.plainweave.requirements_enrichment.v1envelope (previously MCP-only).Fixed
requirements_enrichmentnow dropsrejectedtrace links before building the view — a reviewed-and-rejected binding no longer reads as requirement coverage (present); a rejected-only entity that resolves locally readsabsent.plainweave doctorWardline-findings remediation is now root-aware (cd <root> && wardline scan .when the inspected root is not the cwd).Release steps
main[1.2.0]section + link refv1.2.0at the merge commit🤖 Generated with Claude Code