The mcp_servers key is not a valid key in the Claude plugin manifest
schema. MCP servers are configured in the root .mcp.json file instead.
Affected plugins:
- cmdb-assistant
- contract-validator
- data-platform
- pr-review
- projman
- viz-platform
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Note: This fix may not help because MCP servers fail BEFORE hooks run.
See lessons learned wiki page for full debug trace.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Plugin load failures were caused by version mismatch between
marketplace.json and individual plugin.json files:
- contract-validator: 1.2.0 vs 1.1.0
- git-flow: 1.2.0 vs 1.0.0
- projman: 3.4.0 vs 3.3.0
- clarity-assist: 1.2.0 vs 1.0.0
- doc-guardian: 1.1.0 vs 1.0.0
Claude Code silently fails to load plugins when versions don't match.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Startup hooks in data-platform and pr-review were checking for venvs
at the marketplace path (~/.claude/plugins/marketplaces/.../mcp-servers/)
which gets wiped on updates. The actual venvs live in the cache directory
(~/.cache/claude-mcp-venvs/) which survives updates.
This caused false "MCP venv missing" errors even when venvs existed,
wasting hours of debugging time.
Fixed hooks now check cache path first, matching the pattern used
by run.sh scripts.
Also updated docs/CANONICAL-PATHS.md with the correct venv path pattern
to prevent future occurrences.
Lesson learned: lessons/patterns/startup-hooks-must-check-venv-cache-path-first
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
The startup hooks were looking for MCP venvs relative to the plugin
directory instead of the marketplace root, causing false "venv missing"
errors on every session start.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Add per-agent model selection using Claude Code's now-supported `model`
frontmatter field, and standardize all agent frontmatter across the
marketplace.
Changes:
- Add `model` field to all 25 agents (18 sonnet, 7 haiku)
- Fix viz-platform/data-platform agents using `agent:` instead of `name:`
- Remove non-standard `triggers:` field from domain agents
- Add missing frontmatter to 13 agents
- Document model selection in CLAUDE.md and CONFIGURATION.md
- Fix undocumented commands in README.md
Model assignments based on reasoning depth, tool complexity, and latency:
- sonnet: Planner, Orchestrator, Executor, Coordinator, Security Reviewers
- haiku: Maintainability Auditor, Test Validator, Git Assistant, etc.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Issue 1 - MCP Server Mapping:
- Add mcp_servers field to plugin.json for plugins using shared MCP servers
- projman/pr-review now install gitea MCP server
- cmdb-assistant now installs netbox MCP server
- Scripts read MCP server names from plugin.json
Issue 2 - CLAUDE.md Section Markers:
- Install wraps content with HTML comment markers for precise removal
- Uninstall uses markers first, falls back to legacy header detection
- Fixes code block false positives during uninstall
Bug fix:
- Change ((servers_added++)) to ((++servers_added)) to avoid exit code 1
with set -e when incrementing from 0
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Add 3 new commands for auditing and optimizing Claude Code permission
configurations, leveraging the marketplace's multi-layer review architecture.
New commands:
- /config-audit-settings - 100-point scoring across redundancy, coverage,
safety alignment, and profile fit
- /config-optimize-settings - apply optimizations with dry-run, named
profiles (conservative, reviewed, autonomous), consolidation modes
- /config-permissions-map - Mermaid diagram of review layer coverage
New skill:
- settings-optimization.md - 7 sections covering file formats, syntax
reference, consolidation rules, review-layer-aware recommendations,
named profiles, scoring criteria, and hook detection
Updated agent maintainer.md with new "Audit Settings Files" responsibility.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Added sprint integration section (pre-close verification workflow)
- Added concrete usage examples for all modes
- Added edge cases table
- Added DO NOT rules for both modes
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- progress-tracking.md: new Sprint Dispatch Log section with event types
- orchestrator.md: new responsibility to maintain dispatch log
- Enables timeline reconstruction after interrupted sessions
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- New skill: visual-output.md defines all header, progress, and verdict formats
- All 4 agent files now reference the skill instead of inline templates
- Phase Registry table maps agents to their emoji and phase name
- Single source of truth for visual branding changes
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- design-gate.md and data-gate.md declare gate_contract: v1
- domain-consultation.md Gate Command Reference includes Contract column
- validate_workflow_integration now checks contract version compatibility
- Tests added for match, mismatch, and missing contract scenarios
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- New skill: sprint-lifecycle.md defines states, transitions, and check protocol
- All sprint commands now check and set lifecycle state
- States tracked in milestone description metadata (Sprint/Planning, Sprint/Executing, Sprint/Reviewing)
- Out-of-order calls produce warnings with guidance
- --force override available for all lifecycle checks
- Added Sprint/* labels to label taxonomy documentation
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Created unified /rfc command with create|list|review|approve|reject sub-commands
- Deleted 5 individual rfc-*.md command files
- Moved /clear-cache into /setup --clear-cache
- Updated all cross-references in skills, docs, and integration files
- Command count: 17 -> 12 (net -5)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- sprint-approval.md: approval is now a hard block, not a warning
- sprint-start.md: added --force flag documentation
- orchestrator.md: approval verification is now a hard stop
- docs: updated commands cheatsheet
BREAKING: /sprint-start now requires approval or --force flag
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Domain Advisory Pattern Hardening - patch release to close gaps from v5.6.0/v5.7.0:
## Added
- New `validate_workflow_integration` MCP tool validates domain plugins expose
required advisory interfaces (gate command, review command, advisory agent)
- New `MISSING_INTEGRATION` issue type for workflow integration validation
- New `WorkflowIntegrationResult` Pydantic model for structured validation output
- `integrates_with` field on viz-platform and data-platform in marketplace.json
declaring projman integration metadata
- 4 new test cases for workflow integration validation
## Fixed
- scripts/setup.sh banner version updated from v5.1.0 to v5.7.1
## Documentation
- Updated mcp-tools-reference.md with new tool
- Updated validation-rules.md with Workflow Integration Checks section
- Added /design-gate, /design-review, /data-gate, /data-review to COMMANDS-CHEATSHEET
- Added contract-validator to CONFIGURATION.md plugin table
- Updated README.md Contract Validator tools table
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Comprehensive data integrity audit for human review.
- Detailed report with FAIL/WARN/INFO severity levels
- Schema, lineage, dbt, and PostGIS validation
- Actionable recommendations for each finding
- Graceful degradation when components unavailable
Use cases: sprint planning, code review, post-migration,
periodic health checks, project onboarding.
Closes#376
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Binary pass/fail gate command for projman orchestrator integration.
- Invoked when Domain/Data label present on issue
- Checks FAIL-level violations only (speed optimization)
- Returns compact PASS/FAIL output for automation
- Graceful degradation when database/dbt unavailable
Completes the Domain Advisory Pattern for data domain.
Closes#375
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Create domain-consultation.md skill with detection rules and gate protocols
- Update orchestrator.md to load skill and run domain gates before completion
- Add critical reminder for domain gate enforcement
Closes#356
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Add new Domain category with 2 labels for domain-specific validation gates:
- Domain/Viz: triggers viz-platform design gates for frontend/visualization work
- Domain/Data: triggers data-platform data gates for data engineering work
Update version to 5.6.0 and document Domain Advisory Pattern feature.
Closes#363
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Add standalone command that invokes the design-reviewer agent to
perform detailed design system compliance audits on target paths.
Returns comprehensive findings grouped by severity with file paths,
line numbers, and recommended fixes.
Closes#359
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Add /design-gate command that provides binary pass/fail validation for
design system compliance. This command is used by projman orchestrator
during sprint execution to gate issue completion for Domain/Viz issues.
Features:
- Binary PASS/FAIL output for automation
- Checks only FAIL-level violations (invalid props, missing components)
- Integrates with projman sprint execution workflow
- Lightweight alternative to /design-review
Closes#360
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Implement wiki-based Request for Comments system for capturing,
reviewing, and tracking feature ideas through their lifecycle.
New commands:
- /rfc-create: Create RFC from conversation or clarified spec
- /rfc-list: List RFCs grouped by status
- /rfc-review: Submit Draft RFC for review
- /rfc-approve: Approve RFC for sprint planning
- /rfc-reject: Reject RFC with documented reason
RFC lifecycle: Draft → Review → Approved → Implementing → Implemented
Integration:
- /sprint-plan detects approved RFCs and offers selection
- /sprint-close updates RFC status on completion
- clarity-assist suggests /rfc-create for feature ideas
New MCP tool: allocate_rfc_number
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
The hooks.json files had incorrect structure. Claude Code requires:
- Top-level "hooks" object
- Event names as keys containing arrays
- Each array element must have "matcher" and nested "hooks" array
Fixed 8 plugins:
- clarity-assist
- claude-config-maintainer
- cmdb-assistant
- contract-validator
- data-platform
- pr-review
- projman
- viz-platform
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Three plugins had incorrect hooks.json structure that caused hooks to fail:
- pr-review: SessionStart used nested `hooks` array without matcher
- cmdb-assistant: SessionStart used nested `hooks` array without matcher
- git-flow: Used completely wrong format (array with `event` field)
Per Claude Code documentation:
- Without matcher: direct `type`/`command` in the event array
- With matcher: nested `hooks` array inside matcher object
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Add hooks declarations to 9 plugins missing them in marketplace.json
- Change .mcp.json to use relative paths (portable across users)
- Fix pr-review hook schema to use standard nested hooks structure
- Fix token exposure in cmdb-assistant startup-check.sh (use curl -K)
- Update version to 5.4.1 in marketplace.json, README.md
- Fix CANONICAL-PATHS.md version (was incorrectly showing 5.5.0)
All 12 plugins now have hooks properly registered.
All validations pass.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Extract shared knowledge from 5 command files into 6 reusable skills:
- plugin-discovery.md: Plugin scanning and discovery
- interface-parsing.md: README.md and CLAUDE.md parsing
- dependency-analysis.md: MCP server and data flow analysis
- validation-rules.md: Compatibility and agent validation
- mcp-tools-reference.md: Available MCP tools
- visual-output.md: Standard formatting and headers
Slim commands from 263-164 lines down to 44-55 lines each.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Claude Code rejects `defaultModel` in plugin.json and `model` in agent
frontmatter with "Unrecognized key" validation error.
Removed:
- defaultModel from 6 plugin.json files
- model from 7 agent frontmatter files
- docs/MODEL-RECOMMENDATIONS.md (deleted)
- Model config sections from CONFIGURATION.md and CLAUDE.md
- Model validation from validate-marketplace.sh
This reverts Sprint 7 (v5.4.0) multi-model feature that was never
supported by Claude Code's plugin schema.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
clear-cache.md and suggest-version.md were missing the required YAML
frontmatter with description field. This caused Claude Code to skip
loading the entire projman plugin's commands.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
