docs(#4188): reconcile current→upstream AI-asset mapping tables - #4189
docs(#4188): reconcile current→upstream AI-asset mapping tables#4189fullsend-ai-coder[bot] wants to merge 3 commits into
Conversation
Rewrite mapping tables across all upstream-schema-alignment and ai-catalog-entity-model OpenSpec files to reflect the corrected upstream entity kind targets per Decision 1 and the #4042 gate comment. Key corrections across 7 files: - MCP servers: kind already aligned with upstream (McpServerApiEntity, backstage#34016). Removed all `kind: McpServer` targets and `API → McpServer` examples. Confidence: High. - Skills/rules: `AiResource` shipped upstream (#33575). Updated from "no RFC yet" / Low to Medium-High with kind/name casing alignment as remaining work. - Model server: candidate `API` / `ai-model-server` (#34476 open PR). Updated to Medium/Low with hedge. - AI model: no solid upstream kind. Confidence: Low. - Agent: not attributed to RFC #32062 (MCP-only). Fixed RHIDP-15302 text that incorrectly said #32062 → AIAgent. - Covers all 7 categories (was 5 in migration-readiness). - Updated CLI samples (JSON + human-readable) to show correct field-level transforms instead of kind renames. - Added TBD note for vector-store / ai-tool categories. - Each file cites Decision 1 as current-state SoT. Closes #4188
Codecov Report✅ All modified and coverable lines are covered by tests. Additional details and impacted files@@ Coverage Diff @@
## main #4189 +/- ##
=======================================
Coverage 58.30% 58.30%
=======================================
Files 2427 2427
Lines 96721 96721
Branches 26915 26925 +10
=======================================
Hits 56390 56390
Misses 38869 38869
Partials 1462 1462
*This pull request uses carry forward flags. Click here to find out more. Continue to review full report in Codecov by Harness.
🚀 New features to boost your workflow:
|
|
/fs-fix CI issues |
|
🤖 Finished Fix · ✅ Success · Started 12:39 PM UTC · Completed 12:51 PM UTC |
Run prettier:fix on design.md and proposal.md to fix table column alignment that was failing the CI prettier check. Addresses review feedback on #4189
🔧 Fix agent — iteration 1 (human-triggered)Fixed CI prettier failures in 2 markdown files (design.md and proposal.md) by running prettier:fix to correct table column alignment. Both 'Workspace boost, CI step for node 22' and 'Workspace boost, CI step for node 24' were failing on the prettier check step. Fixed (2):
Tests: passed Updated by fullsend fix agent |
|
/fs-review |
|
🤖 Finished Review · ✅ Success · Started 1:23 PM UTC · Completed 1:39 PM UTC |
ReviewFindingsMedium
Low
Labels: All changes are documentation files under workspaces/boost/openspec/, matching the workspace/boost and documentation labels. |
![fullsend-ai-review[bot]](https://avatars.githubusercontent.com/in/3478501?s=60&v=4){kind=small}
There was a problem hiding this comment.
See the review comment for full details.
Note: The following inline comments could not be posted on the diff (GitHub returned 422) and are included here instead:
workspaces/boost/openspec/changes/ai-catalog-entity-model/design.md(file-level): Line 75 · [medium] internal-inconsistency
The PR updates the Context paragraph to state 'Agent and AI model kinds are not yet proposed upstream,' but the Migration path example in the same file still references kind: AIAgent as a concrete target. This contradicts the PR's correction that no upstream agent kind exists.
Suggested fix: Update the Migration path example to use a different illustration (e.g., AiResource casing alignment for skills) or add a qualifier noting that AIAgent is hypothetical.
workspaces/boost/openspec/changes/upstream-schema-alignment/design.md:56: [medium] stale-reference
Decision 2's bullet-list description still says 'Proposed RFC kind' and 'Confidence level (high/medium/low),' but the PR updated the table header to 'Upstream Target' and the confidence definitions to a four-tier system. The bullets were not updated to match.
Suggested fix: Change 'Proposed RFC kind' to 'Upstream target kind' and update the confidence level list to match the four-tier taxonomy.
workspaces/boost/openspec/changes/upstream-schema-alignment/tasks.md(file-level): Line 54 · [medium] stale-reference
Tasks 4.7 and 5.3 still reference 'RFC #33060,' which the PR's own changes (tasks 1.5, 1.6, 1.6a, 1.6b) replace with specific upstream targets (#34016, #33575, #34476).
Suggested fix: Update tasks 4.7 and 5.3 to reference the current upstream targets: #34016 (McpServerApiEntity), #33575 (AiResource), #34476 (model-server candidate).
workspaces/boost/openspec/changes/upstream-schema-alignment/design.md(file-level): Line 491 · [low] stale-reference
Decision 5 prescribes the dry-run tool footer as 'pending RFC finalization' but the tooling spec output was updated to 'pending upstream stabilization.'
Suggested fix: Update Decision 5 footer text to 'pending upstream stabilization.'
workspaces/boost/openspec/changes/ai-catalog-entity-model/specs/migration-readiness/spec.md:68: [low] edge-case
The 'Entity refs change format' scenario shows identical WHEN/THEN entity refs (airesource:default/my-skill). Backstage uses lowercase kind prefixes, so AIResource to AiResource does not change the ref string. The scenario title is misleading for this example.
fullsend-ai-review
Bot
added
requires-manual-review
|
/fs-fix per review comments so far |
|
🤖 Finished Fix · ✅ Success · Started 1:43 PM UTC · Completed 2:00 PM UTC |
- Replace hypothetical AIAgent migration example with actual AiResource casing alignment example in ai-catalog-entity-model/design.md - Update Decision 2 bullet list to say "Upstream target kind" and four-tier confidence in upstream-schema-alignment/design.md - Replace stale RFC #33060 references in tasks 4.7 and 5.3 with current upstream targets (#34016, #33575, #34476) - Update annotation-scheme/spec.md example to use AiResource mapping - Replace RFC #33060 reference in ai-catalog-entity-model/proposal.md - Align Decision 5 footer text to "pending upstream stabilization" - Fix entity refs scenario to clarify no ref change for casing-only alignment (Backstage lowercases kind prefixes) - Standardize en-dash in human-readable Medium–High confidence label - Update RFC #33060 references in boost-frontend-architecture.md and ai-catalog-frontend/design.md Addresses review feedback on #4189
🔧 Fix agent — iteration 2 (human-triggered)Addressed all 10 review findings (3 medium, 7 low). Fixed stale RFC #33060 references across 5 files, resolved internal inconsistency with AIAgent migration example, aligned Decision 2 bullet list and Decision 5 footer text with current content, clarified entity refs scenario for casing-only changes, and standardized en-dash usage in confidence labels. Fixed (10):
Tests: passed Updated by fullsend fix agent |
|
1 participant
Rewrite mapping tables across all upstream-schema-alignment and ai-catalog-entity-model OpenSpec files to reflect the corrected upstream entity kind targets per Decision 1 and the #4042 gate comment.
Key corrections across 7 files:
(McpServerApiEntity, backstage#34016). Removed all
kind: McpServertargets andAPI → McpServerexamples.Confidence: High.
AiResourceshipped upstream (#33575).Updated from "no RFC yet" / Low to Medium-High with
kind/name casing alignment as remaining work.
API/ai-model-server(#34476 open PR). Updated to Medium/Low with hedge.
RHIDP-15302 text that incorrectly said #32062 → AIAgent.
correct field-level transforms instead of kind renames.
Closes #4188
Post-script verification
agent/4188-reconcile-mapping-tables)4a8aa72ec6e2e1a2158ad6a31354455643b7a573..HEAD)