feat(boost): RHDHPLAN-1513 openspecs - #3755
Conversation
…m alignment Create OpenSpec changes covering all 4 epics and 13 stories under RHDHPLAN-1513 (AI Catalog Ingestion Operations & Upstream Schema Alignment): - ingestion-health-dashboard (RHIDP-15331): 5 stories, 3 specs, 7 files Health status API, admin health UI, error classification - connector-config-hot-reload (RHIDP-15332): 3 stories, 3 specs, 7 files Zod config schemas, RuntimeConfigResolver hot-reload, config admin UI - ingestion-audit-metrics (RHIDP-15333): 3 stories, 3 specs, 7 files Sync audit events, analytics REST API, Eval Hub integration - upstream-schema-alignment (RHIDP-15334): 2 stories, 2 specs, 6 files Annotation specification document, dry-run migration-readiness tooling Cross-references RHDHPLAN-1507 (entity model), RHDHPLAN-1508 (audit logging pattern), RHDHPLAN-1510 (connectors), and upstream Backstage RFCs #32062 and #33060. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> Signed-off-by: gabemontero <gmontero@redhat.com>
RHIDP-15333 (Ingestion Audit Logging & Metrics) closed and absorbed by RHIDP-15277 (AI Catalog RBAC Audit Logging, RHDHPLAN-1508). Closed epic openspec (ingestion-audit-metrics/): - .openspec.yaml status → consolidated with absorbing epic metadata - proposal.md, design.md, tasks.md: consolidation headers with story mapping - specs/audit-events: status → Consolidated into RHIDP-15280 - specs/analytics-api: status → Consolidated into RHIDP-15277 - specs/eval-hub-integration: status → Consolidated into RHIDP-15277 Surviving epic openspecs (consolidation cross-references): - ingestion-health-dashboard: proposal.md + design.md annotations - connector-config-hot-reload: proposal.md + design.md annotations - upstream-schema-alignment: proposal.md + design.md annotations Jira updates: - RHIDP-15343 → CONSOLIDATED into RHIDP-15280, closed as Duplicate - RHIDP-15344 → CONSOLIDATED into RHIDP-15277 scope, closed as Duplicate - RHIDP-15345 → CONSOLIDATED into RHIDP-15277 scope, closed as Duplicate - Blocks links: RHIDP-15295→15338, RHIDP-15258→15346, RHIDP-15258→15347 Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> Signed-off-by: gabemontero <gmontero@redhat.com>
|
🤖 Finished Review · ✅ Success · Started 7:51 PM UTC · Completed 8:02 PM UTC |
PR Summary by QodoAdd Boost OpenSpecs for RHDHPLAN-1513 epics (with audit-metrics consolidation)
AI Description
Diagram
High-Level Assessment
Files changed (27)
|
Codecov Report✅ All modified and coverable lines are covered by tests. Additional details and impacted files@@ Coverage Diff @@
## main #3755 +/- ##
==========================================
+ Coverage 54.65% 54.68% +0.02%
==========================================
Files 2360 2362 +2
Lines 90140 90156 +16
Branches 25214 25214
==========================================
+ Hits 49270 49302 +32
- Misses 40598 40634 +36
+ Partials 272 220 -52
*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:
|
ReviewFindingsMedium
Low
Previous runReviewFindingsMedium
Low
Previous run (2)ReviewFindingsMedium
Low
Previous run (3)ReviewFindingsMedium
Low
Previous run (4)ReviewFindingsLow
Previous run (5)ReviewFindingsMedium
Low
Previous run (6)ReviewFindingsMedium
Low
Previous run (7)ReviewFindingsMedium
Low
Previous run (8)ReviewFindingsMedium
Low
Previous run (9)ReviewFindingsMedium
Low
Previous run (10)ReviewFindingsMedium
Low
Previous run (11)ReviewFindingsMedium
Low
Previous run (12)ReviewVerdict: Comment — medium-severity findings worth noting for spec accuracy, but none blocking for draft pre-implementation specifications. SummaryThis PR adds 27 new openspec documents (design, proposal, specs, tasks) for four RHDHPLAN-1513 feature areas under However, several specifications describe API surfaces that diverge from the existing codebase without acknowledging the gap. Since these specs describe "extending existing infrastructure," clarity about what's reused vs. new is important for implementation accuracy. FindingsMedium1. API surface assumptions diverge from existing The specs reference methods that don't exist on the current classes:
The designs claim to "extend" and "reuse" existing infrastructure, but silently introduce new API surface. This gap could mislead implementers about the scope of changes needed. Remediation: Add a section to 2. The specs assume Remediation: Document in the design that the 3. Permission name inconsistency across feature areas ( The Remediation: Standardize on 4. Health status "degraded" definition inconsistency ( The proposal.md defines degraded as "last sync succeeded but recent failures" (most recent must have succeeded). The design.md defines it as "Mixed results (not all-success and not all-failure)" (broader — includes cases where most recent failed). The Remediation: Align Low5. "EXISTING Requirements" label applied to non-existent features ( Several spec files label their scenarios as "EXISTING Requirements" but describe behavior for Jira, GitHub, and GitLab connectors that don't exist in the codebase. The entity providers that exist are 6. Misleading worst-case propagation timing ( The design states worst-case latency as "Cache TTL (30s) + reconciliation interval (5m) = 5m30s". But the same document describes immediate cache invalidation on DB write. The 30s TTL only applies as a safety net if invalidation fails. Normal-case latency is just the reconciliation interval (~5m). 7. SQL The 8. The design.md lists three Dimensions Reviewed
Labels: PR modifies only the boost workspace openspec documentation directory. Previous run (13)ReviewVerdict: Approve This PR adds OpenSpec documentation for four feature areas under RHDHPLAN-1513: Connector Configuration Hot-Reload, Ingestion Audit Logging & Metrics, Ingestion Health Admin Dashboard, and Upstream Schema Alignment Readiness. All 27 files are documentation-only (markdown specifications and YAML metadata) with no code changes. AssessmentStructure & conventions: All four feature areas follow the established OpenSpec pattern already in the repository ( Consolidation handling: The Cross-references: RHDHPLAN and RHIDP ticket references are consistent throughout. The consolidation notes in each design/proposal document correctly describe which stories migrated where, and surviving epics are identified. Spec quality: The specifications are thorough — design docs include clear decisions with rationale, scenarios use GIVEN/WHEN/THEN format, and task breakdowns include JIRA story references and priority levels. The connector-config-hot-reload design, for example, documents the configScope annotation strategy with a clear rationale for which fields are Low-severity finding
SummaryThis is a well-structured documentation PR that follows established repository conventions. The specifications provide comprehensive coverage of four feature areas with clear design decisions, rationale, and task breakdowns. The one minor inconsistency noted above does not block approval. Labels: PR adds OpenSpec documentation for boost workspace ingestion features |
fullsend-ai-review
Bot
added
the
requires-manual-review
Code Review by Qodo
Context used✅ Compliance rules (platform):
37 rules✅ Cross-repo context Not relevant to this PR:
redhat-developer/rhdh Not relevant to this PR:
redhat-developer/rhdh-chart Not relevant to this PR:
redhat-developer/rhdh-operator Not relevant to this PR:
redhat-developer/rhdh-local 1.
|
rhdh-qodo-merge
Bot
added
documentation
|
Thanks for the thorough review. Addressing all 9 findings: 1. Credential rotation latency mismatch — Agreed. The spec correctly calculates ~6 minutes (60s kubelet + 5min reconciliation). The ≤90s claim in 2. API endpoint pattern inconsistency — Agreed. Standardizing all admin config save calls on 3. Disabled connector filtering conflict — Agreed. Updating 4. PostgreSQL-incompatible INDEX syntax — Agreed. Replacing MySQL-style inline 5. Undefined health state — Agreed. Redefining health derivation: Healthy = all N succeeded, Degraded = mixed (not all-success and not all-failure), Failing = all N failed. This covers all combinations including [success, success, failure]. 6. Underspecified edge cases — Agreed. Adding explicit derivation for fewer-than-3 attempts: 0 = Unknown (never synced), 1+ all-success = Healthy, all-failure = Failing, mixed = Degraded. 7. Missing 8. Orphaned 9. Inconsistent UI propagation messages — Agreed. Updating "within 30 seconds" to "within 30 seconds + next reconciliation cycle" for consistency with the propagation spec. |
Rebase consistency fix: update stale skill entity references to match confirmed category taxonomy (skill → AIResource kind, spec.type: skill). - proposal.md: Skill row Resource/ai-skill → AIResource/skill - annotation-specification/spec.md: ai-skill (Resource kind) → skill (AIResource kind) - ingestion-audit-metrics/design.md: spec.type: ai-skill → spec.type: skill (2 locations) Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
…nspecs Addresses all 9 findings from fullsend review agent on PR redhat-developer#3755: 1. Credential rotation latency: fix ≤90s claim → ~6 minutes (60s kubelet + 5min reconciliation) in design.md Decision 4 and Risks section 2. API endpoint pattern: standardize on POST /api/boost/admin/config with key-value body; fix missing /api/boost prefix in toggle scenarios 3. Disabled connector filtering: reconcile design.md with spec — excluded by default, available via ?includeDisabled=true 4. PostgreSQL INDEX syntax: replace MySQL-style inline INDEX with separate CREATE INDEX statements in both design.md files 5. Health derivation gap: redefine as Healthy/Degraded/Failing/Unknown covering all attempt combinations including [success, success, failure] 6. Fewer-than-3 edge cases: explicit derivation for 0-2 attempts using same rules as 3-attempt case 7. Missing tls.caFile: add to Jira connector config schema in spec 8. Orphaned db-only configScope: remove from proposal.md 9. Propagation messages: fix 'within 30 seconds' → 'within 30 seconds + next reconciliation cycle' Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> Signed-off-by: gabemontero <gmontero@redhat.com>
|
🤖 Review · ❌ Terminated · Started 8:41 PM UTC · Ended 8:47 PM UTC |
![fullsend-ai-review[bot]](https://avatars.githubusercontent.com/in/3478501?s=60&v=4){kind=small}
This comment was marked as outdated.
This comment was marked as outdated.
Sorry, something went wrong.
fullsend-ai-review
Bot
added
ready-for-merge
|
🤖 Finished Review · ✅ Success · Started 8:41 PM UTC · Completed 8:47 PM UTC |
…ly state Re-add db-only to proposal.md configScope list (reverses earlier removal). Add concrete db-only fields (lastSyncTimestamp, lastSyncOutcome, runStatus) to design.md Decision 2 table with rationale paragraph explaining these are pure runtime state with no YAML baseline. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> Signed-off-by: gabemontero <gmontero@redhat.com>
|
🤖 Finished Review · ✅ Success · Started 10:00 PM UTC · Completed 10:13 PM UTC |
fullsend-ai-review
Bot
added
requires-manual-review
mareklibra
left a comment
mareklibra
left a comment
There was a problem hiding this comment.
A few comments added - mostly clarifications or conflicts in specs.
…nspecs Addresses all 9 findings from fullsend review agent on PR redhat-developer#3755: 1. Credential rotation latency: fix ≤90s claim → ~6 minutes (60s kubelet + 5min reconciliation) in design.md Decision 4 and Risks section 2. API endpoint pattern: standardize on POST /api/boost/admin/config with key-value body; fix missing /api/boost prefix in toggle scenarios 3. Disabled connector filtering: reconcile design.md with spec — excluded by default, available via ?includeDisabled=true 4. PostgreSQL INDEX syntax: replace MySQL-style inline INDEX with separate CREATE INDEX statements in both design.md files 5. Health derivation gap: redefine as Healthy/Degraded/Failing/Unknown covering all attempt combinations including [success, success, failure] 6. Fewer-than-3 edge cases: explicit derivation for 0-2 attempts using same rules as 3-attempt case 7. Missing tls.caFile: add to Jira connector config schema in spec 8. Orphaned db-only configScope: remove from proposal.md 9. Propagation messages: fix 'within 30 seconds' → 'within 30 seconds + next reconciliation cycle' Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> Signed-off-by: gabemontero <gmontero@redhat.com>
- Fix augment workspace paths to boost paths in proposal.md Impact - Standardize config keys to boost.connectors.* namespace (BoostConfigKey) - Align method names: resolve(), invalidate() (no args), setOverride() - Align tasks.md REST endpoints with design.md POST /api/boost/admin/config - Document flat-key write semantics (single leaf BoostConfigKey per write) - Clarify startup gate (catalog.providers.*.enabled) vs runtime sync-skip - Align degraded health derivation with design.md (mixed results) - Remove db-only runtime state from configScope table (health store owns it) - Replace in-memory run registry with coreServices.cache/SchedulerService Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> Signed-off-by: gabemontero <gmontero@redhat.com>
…4j field alignment - error-classification/spec.md: error_type → errorType in THEN clauses (14 occurrences). DB column is snake_case error_type, but ErrorClassifier returns camelCase errorType per the interface at line 105 and tasks.md. - analytics-api/spec.md: neo4j_status.connected → neo4j_connected to match design.md Decision 2 line 113 flat field structure. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> Signed-off-by: gabemontero <gmontero@redhat.com>
Cascade propagation audit — 2 fixes (
|
|
Grill-me results: |
…te model extends, not matches, Decision 1 enum Decision 1 defines 4 health status values (healthy/degraded/failing/unknown). Decision 4's five-state model adds Disabled as an enabled=false overlay — it extends the enum, not matches it. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> Signed-off-by: gabemontero <gmontero@redhat.com>
|
🤖 Review · ❌ Terminated · Started 10:48 PM UTC · Ended 11:09 PM UTC |
|
🤖 Finished Review · ✅ Success · Started 10:48 PM UTC · Completed 11:08 PM UTC |
What GitHub is flagging is mareklibra's "Changes Requested" review (submitted July 20, review ID 4733251845). Even though all of mareklibra's inline comments have been addressed and resolved, the review itself still has state CHANGES_REQUESTED. Mareklibra's follow-up on July 22 was only a COMMENTED review — it didn't clear the "Changes Requested" state.
gabemontero
added
ready-for-merge
mareklibra
left a comment
mareklibra
left a comment
There was a problem hiding this comment.
Based on the previous comments, I am not flagging ai-catalog.admin and boost.admin any further. However, we should align them before implementation.
…nspecs Addresses all 9 findings from fullsend review agent on PR redhat-developer#3755: 1. Credential rotation latency: fix ≤90s claim → ~6 minutes (60s kubelet + 5min reconciliation) in design.md Decision 4 and Risks section 2. API endpoint pattern: standardize on POST /api/boost/admin/config with key-value body; fix missing /api/boost prefix in toggle scenarios 3. Disabled connector filtering: reconcile design.md with spec — excluded by default, available via ?includeDisabled=true 4. PostgreSQL INDEX syntax: replace MySQL-style inline INDEX with separate CREATE INDEX statements in both design.md files 5. Health derivation gap: redefine as Healthy/Degraded/Failing/Unknown covering all attempt combinations including [success, success, failure] 6. Fewer-than-3 edge cases: explicit derivation for 0-2 attempts using same rules as 3-attempt case 7. Missing tls.caFile: add to Jira connector config schema in spec 8. Orphaned db-only configScope: remove from proposal.md 9. Propagation messages: fix 'within 30 seconds' → 'within 30 seconds + next reconciliation cycle' Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> Signed-off-by: gabemontero <gmontero@redhat.com>
…ng, removeOverride Thread 1: catalog.providers → ai-catalog.providers across 7 files (11 edits) Threads 2-4: 'within 30s' → 'cache refresh ≤30s; takes effect on next reconciliation cycle' across 5 files (9 edits) Thread 5: add removeOverride/DELETE mechanism for clearing DB overrides, specify cron-over-intervalMs precedence, add matching UI/tasks/schema specs Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> Signed-off-by: gabemontero <gmontero@redhat.com>
|
🤖 Review · ❌ Terminated · Started 7:33 PM UTC · Ended 7:50 PM UTC |
This comment was marked as outdated.
This comment was marked as outdated.
Sorry, something went wrong.
fullsend-ai-review
Bot
added
requires-manual-review
|
🤖 Finished Review · ✅ Success · Started 7:33 PM UTC · Completed 7:50 PM UTC |
|
Re: fullsend review (4790936057) — all findings are pre-existing, unrelated to Marek's review fixes The commit addressing @mareklibra 's 5 review threads was Fixed 5 of 6 in
Not fixing: | 6 | The remaining summary-only findings (spec format deviations in |
1. Confluence → GitLab typo in ingestion-audit-metrics/tasks.md 2. runStatus clarified as derived, not a stored column (design.md) 3. TIMESTAMP → TIMESTAMPTZ in boost_quality_scores schema 4. __schemaVersion Zod/boostConfigFields interaction clarified 5. Cron example backtick-wrapped to prevent markdown mangling All pre-existing issues unrelated to the Marek review fixes in a1edca3c07. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> Signed-off-by: gabemontero <gmontero@redhat.com>
|
🤖 Review · ❌ Terminated · Started 8:00 PM UTC · Ended 8:19 PM UTC |
|
|
🤖 Finished Review · ✅ Success · Started 8:00 PM UTC · Completed 8:19 PM UTC |
|
ok @mareklibra both your comments and fullsend-review's have been processed PTAL |
* chore(boost): RHDHPLAN-1513 openspec content for operations & upstream alignment Create OpenSpec changes covering all 4 epics and 13 stories under RHDHPLAN-1513 (AI Catalog Ingestion Operations & Upstream Schema Alignment): - ingestion-health-dashboard (RHIDP-15331): 5 stories, 3 specs, 7 files Health status API, admin health UI, error classification - connector-config-hot-reload (RHIDP-15332): 3 stories, 3 specs, 7 files Zod config schemas, RuntimeConfigResolver hot-reload, config admin UI - ingestion-audit-metrics (RHIDP-15333): 3 stories, 3 specs, 7 files Sync audit events, analytics REST API, Eval Hub integration - upstream-schema-alignment (RHIDP-15334): 2 stories, 2 specs, 6 files Annotation specification document, dry-run migration-readiness tooling Cross-references RHDHPLAN-1507 (entity model), RHDHPLAN-1508 (audit logging pattern), RHDHPLAN-1510 (connectors), and upstream Backstage RFCs #32062 and #33060. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> Signed-off-by: gabemontero <gmontero@redhat.com> * fix(boost): RHDHPLAN-1513 openspec consolidation for RHIDP-15333 closure RHIDP-15333 (Ingestion Audit Logging & Metrics) closed and absorbed by RHIDP-15277 (AI Catalog RBAC Audit Logging, RHDHPLAN-1508). Closed epic openspec (ingestion-audit-metrics/): - .openspec.yaml status → consolidated with absorbing epic metadata - proposal.md, design.md, tasks.md: consolidation headers with story mapping - specs/audit-events: status → Consolidated into RHIDP-15280 - specs/analytics-api: status → Consolidated into RHIDP-15277 - specs/eval-hub-integration: status → Consolidated into RHIDP-15277 Surviving epic openspecs (consolidation cross-references): - ingestion-health-dashboard: proposal.md + design.md annotations - connector-config-hot-reload: proposal.md + design.md annotations - upstream-schema-alignment: proposal.md + design.md annotations Jira updates: - RHIDP-15343 → CONSOLIDATED into RHIDP-15280, closed as Duplicate - RHIDP-15344 → CONSOLIDATED into RHIDP-15277 scope, closed as Duplicate - RHIDP-15345 → CONSOLIDATED into RHIDP-15277 scope, closed as Duplicate - Blocks links: RHIDP-15295→15338, RHIDP-15258→15346, RHIDP-15258→15347 Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> Signed-off-by: gabemontero <gmontero@redhat.com> * fix(boost): RHDHPLAN-1513 align skill entity mapping to AIResource/skill Rebase consistency fix: update stale skill entity references to match confirmed category taxonomy (skill → AIResource kind, spec.type: skill). - proposal.md: Skill row Resource/ai-skill → AIResource/skill - annotation-specification/spec.md: ai-skill (Resource kind) → skill (AIResource kind) - ingestion-audit-metrics/design.md: spec.type: ai-skill → spec.type: skill (2 locations) Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix(boost): RHDHPLAN-1513 address fullsend review findings across openspecs Addresses all 9 findings from fullsend review agent on PR redhat-developer#3755: 1. Credential rotation latency: fix ≤90s claim → ~6 minutes (60s kubelet + 5min reconciliation) in design.md Decision 4 and Risks section 2. API endpoint pattern: standardize on POST /api/boost/admin/config with key-value body; fix missing /api/boost prefix in toggle scenarios 3. Disabled connector filtering: reconcile design.md with spec — excluded by default, available via ?includeDisabled=true 4. PostgreSQL INDEX syntax: replace MySQL-style inline INDEX with separate CREATE INDEX statements in both design.md files 5. Health derivation gap: redefine as Healthy/Degraded/Failing/Unknown covering all attempt combinations including [success, success, failure] 6. Fewer-than-3 edge cases: explicit derivation for 0-2 attempts using same rules as 3-attempt case 7. Missing tls.caFile: add to Jira connector config schema in spec 8. Orphaned db-only configScope: remove from proposal.md 9. Propagation messages: fix 'within 30 seconds' → 'within 30 seconds + next reconciliation cycle' Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> Signed-off-by: gabemontero <gmontero@redhat.com> * fix(boost): RHDHPLAN-1513 document db-only configScope for runtime-only state Re-add db-only to proposal.md configScope list (reverses earlier removal). Add concrete db-only fields (lastSyncTimestamp, lastSyncOutcome, runStatus) to design.md Decision 2 table with rationale paragraph explaining these are pure runtime state with no YAML baseline. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> Signed-off-by: gabemontero <gmontero@redhat.com> * fix(boost): RHDHPLAN-1513 address review feedback on openspec docs - Fix augment workspace paths to boost paths in proposal.md Impact - Standardize config keys to boost.connectors.* namespace (BoostConfigKey) - Align method names: resolve(), invalidate() (no args), setOverride() - Align tasks.md REST endpoints with design.md POST /api/boost/admin/config - Document flat-key write semantics (single leaf BoostConfigKey per write) - Clarify startup gate (catalog.providers.*.enabled) vs runtime sync-skip - Align degraded health derivation with design.md (mixed results) - Remove db-only runtime state from configScope table (health store owns it) - Replace in-memory run registry with coreServices.cache/SchedulerService Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> Signed-off-by: gabemontero <gmontero@redhat.com> * fix(boost): RHDHPLAN-1513 address fullsend review findings on PR redhat-developer#3755 Fix 1 (medium): getConfig() → resolve(), connectors.jira → boost.connectors.jira in config-schemas/spec.md — missed in Marek fix Fix 2: Remove stale db-only configScope from proposal.md and tasks.md — runtime state lives in health store, not config resolver (per Marek) Fix 3: Add tls.caFile to task 1.1 Jira Zod schema field list — matches design.md configScope table Fix 4: Replace :connectorId Express-style placeholder with <connectorId> in task 6.1 query param Fix 5: Add CHECK(score >= 0.0 AND score <= 1.0) constraint to boost_quality_scores DECIMAL column in ingestion-audit-metrics design Fix 6: Remove unconventional Spec: title prefix from 5 spec files Fix 7: Standardize status header format in 2 upstream-schema-alignment specs — bold+emoji → blockquote format matching other specs Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> Signed-off-by: gabemontero <gmontero@redhat.com> * fix(boost): RHDHPLAN-1513 address Marek review feedback on PR redhat-developer#3755 Thread 1: augment → boost workspace references in ingestion-health-dashboard Thread 2: resolve() leaf key alignment — use boost.connectors.jira.enabled not boost.connectors.jira (namespace) Thread 3: drop 'in-memory or' from task 6.2 — SchedulerService only Thread 4: sync_attempts → boost_sync_attempts table naming convention Thread 6: add config namespace mapping table, remove tls.caFile from boost.connectors scope (belongs under catalog.providers) Thread 7: add unknown status to health status enum in proposal and spec Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> Signed-off-by: gabemontero <gmontero@redhat.com> * fix(boost): RHDHPLAN-1513 address fullsend review findings on PR redhat-developer#3755 - ingestion-audit-metrics/design.md: add permission namespace rationale blockquote explaining ai-catalog.admin vs boost.admin coexistence - ingestion-health-dashboard/design.md: update Decision 4 from three-state to five-state model, adding Degraded and Unknown UX states to match Decision 1 health status enum Assisted-by: claude-opus-4-6 <noreply@anthropic.com> Signed-off-by: gabemontero <gmontero@redhat.com> * fix(boost): RHDHPLAN-1513 address mareklibra review — permission namespace, schema cleanup, enum alignment - Realign permission namespace: boost.admin → ai-catalog.admin across ingestion-health-dashboard and connector-config-hot-reload specs (RHDHPLAN-1513 is catalog ingestion operations, not agent operations) - Replace requireBoostAdmin() middleware with permissions.authorize() using ai-catalog.admin permission per AGENTS.md framework requirement - Fix health status enum: four-state → five-state model (add degraded), add Unknown to status badge colors and task lists - Fix bare connectors.jira.* → boost.connectors.jira.* in hot-reload-propagation spec and tasks - Remove tls.caFile, credentials.*, namespace from boost.connectors Zod schemas (these are catalog.providers fields, not boost.connectors) - Clarify dual enabled gates: catalog.providers.*.enabled (startup registration) vs boost.connectors.*.enabled (runtime sync-skip) - Fix admin_config → boost_admin_config table name - Clarify schema versioning with leaf-only DB writes and connector-level Zod validation - Fix package name @boost/migration-readiness → @red-hat-developer-hub/backstage-plugin-boost-migration-readiness Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> Signed-off-by: gabemontero <gmontero@redhat.com> * fix(boost): RHDHPLAN-1513 address fullsend review — entity mapping, cross-ref, field subset - Model Server mapping: Component/model-server → Resource/ai-model-server (aligns with ai-catalog-entity-model/design.md Decision 1 on main) - Add note: will pivot to upstream Backstage API extension (backstage/backstage#34476) if it becomes available - Cross-reference: catalog-entities/spec.md → ai-catalog-entity-model/design.md Decision 1 (covers all 7 category values) - Tasks 1.3/1.4: 'same field pattern' → connector-appropriate field subset (GitHub/GitLab have 4 fields vs Jira's 6) Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> Signed-off-by: gabemontero <gmontero@redhat.com> * fix(boost): RHDHPLAN-1513 add rule and skill-bundle to annotation values and mapping table Align annotation-specification/spec.md and proposal.md with ai-catalog-entity-model/design.md Decision 1 which defines 7 category values. The annotation listing and mapping table previously listed only 5, omitting rule and skill-bundle. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> Signed-off-by: gabemontero <gmontero@redhat.com> * fix(boost): RHDHPLAN-1513 complete 5-to-7 alignment in tasks.md and spec.type list Continue alignment with ai-catalog-entity-model/design.md Decision 1: - tasks.md task 1.1: add rule, skill-bundle to category values - tasks.md task 3.1: 'five' -> 'seven' AI Asset types - annotation-specification/spec.md: add rule, skill-bundle to spec.type list; fix model-server -> ai-model-server spec.type Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> Signed-off-by: gabemontero <gmontero@redhat.com> * fix(boost): RHDHPLAN-1513 address mareklibra review round — package name, status enum, config schema clarity - Replace @boost/migration-readiness with @red-hat-developer-hub/backstage-plugin-boost-migration-readiness across tasks.md, design.md, and spec.md (4 locations) + add workspace path - Remove tls/credentials/yaml-only from connector-config-hot-reload/proposal.md (those fields live under catalog.providers, not boost.connectors) - Fix health status enum: remove disabled from status values (healthy/degraded/failing/unknown); disabled is an enabled=false overlay, not a status value — propagated to design.md and tasks.md - Clarify per-leaf Zod validation vs connector-level aggregate validation in config-schemas/spec.md - Clarify schemaVersion storage: explicit __schemaVersion leaf key, bump on migration only, concurrent leaf writes non-conflicting, GET-prefix merge behavior Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> Signed-off-by: gabemontero <gmontero@redhat.com> * fix(boost): RHDHPLAN-1513 align spec.type skill-bundle → ai-skill-bundle per entity model proposal.md mapping table and annotation-specification spec.type list both had spec.type: skill-bundle — the authoritative entity model design (ai-catalog-entity-model/design.md Decision 1, on main) defines this as ai-skill-bundle. The rhdh.io/ai-asset-category annotation value remains skill-bundle (correct — category ≠ spec.type). Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> Signed-off-by: gabemontero <gmontero@redhat.com> * fix(boost): RHDHPLAN-1513 cascade propagation — errorType casing, neo4j field alignment - error-classification/spec.md: error_type → errorType in THEN clauses (14 occurrences). DB column is snake_case error_type, but ErrorClassifier returns camelCase errorType per the interface at line 105 and tasks.md. - analytics-api/spec.md: neo4j_status.connected → neo4j_connected to match design.md Decision 2 line 113 flat field structure. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> Signed-off-by: gabemontero <gmontero@redhat.com> * fix(boost): RHDHPLAN-1513 correct Decision 4 parenthetical — five-state model extends, not matches, Decision 1 enum Decision 1 defines 4 health status values (healthy/degraded/failing/unknown). Decision 4's five-state model adds Disabled as an enabled=false overlay — it extends the enum, not matches it. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> Signed-off-by: gabemontero <gmontero@redhat.com> * fix(boost): RHDHPLAN-1513 address Marek review — namespace, TTL wording, removeOverride Thread 1: catalog.providers → ai-catalog.providers across 7 files (11 edits) Threads 2-4: 'within 30s' → 'cache refresh ≤30s; takes effect on next reconciliation cycle' across 5 files (9 edits) Thread 5: add removeOverride/DELETE mechanism for clearing DB overrides, specify cron-over-intervalMs precedence, add matching UI/tasks/schema specs Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> Signed-off-by: gabemontero <gmontero@redhat.com> * fix(boost): RHDHPLAN-1513 address fullsend review — pre-existing issues 1. Confluence → GitLab typo in ingestion-audit-metrics/tasks.md 2. runStatus clarified as derived, not a stored column (design.md) 3. TIMESTAMP → TIMESTAMPTZ in boost_quality_scores schema 4. __schemaVersion Zod/boostConfigFields interaction clarified 5. Cron example backtick-wrapped to prevent markdown mangling All pre-existing issues unrelated to the Marek review fixes in a1edca3c07. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> Signed-off-by: gabemontero <gmontero@redhat.com> --------- Signed-off-by: gabemontero <gmontero@redhat.com> Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
2 participants
Hey, I just made a Pull Request!
These are the openspec markdowns through tasks.md for all the RHIDP epics / stories under RHDHPLAN-1513
✔️ Checklist