Two misalignments found and fixed:
## 1. Pagination response field naming
Ring standards previously documented cursor fields as camelCase
(nextCursor, prevCursor, hasMore). The actual Midaz Pagination struct
uses snake_case json tags:
NextCursor string `json:"next_cursor,omitempty"`
PrevCursor string `json:"prev_cursor,omitempty"`
There is no HasMore/hasMore field in Midaz.
Changes:
- api-patterns.md: fix struct examples, JSON examples, detection
commands, tables, and anti-rationalization rows
- golang.md (opencode): pagination section was already correct
- PM skill files: update pagination field reference tables
## 2. BusinessError struct missing title field
Ring's BusinessError lacked the title field present in both
midaz/pkg/mmodel/error.go and lib-commons ErrorResponse:
type Error struct {
Code string `json:"code"`
Title string `json:"title"`
Message string `json:"message"`
EntityType string `json:"entityType,omitempty"`
}
Changes:
- platforms/opencode/standards/golang.md: add Title field to
BusinessError struct and to ErrInvalidInput/ErrNotFound examples
Files changed:
- dev-team/docs/standards/golang/api-patterns.md
- platforms/opencode/standards/golang.md
- pm-team/skills/pre-dev-api-design/SKILL.md
- pm-team/skills/shared-patterns/standards-discovery.md
Fix QueueMessage struct tags in messaging.md and opencode/golang.md (organization_id → organizationId, etc.) and invert GOOD/BAD reference examples in production-readiness-audit to match the mandatory camelCase JSON convention from api-patterns.md.
X-Lerian-Ref: 0x1
Replace all 13 panic() positive examples with return error patterns, close the bootstrap exception loophole, resolve os.Exit() contradiction in domain.md, fix InitServers() caller to handle (*Service, error), update License Manager guidance to return error instead of panic, make grep commands recursive with regexp.MustCompile exclusion, and add post-generation self-checks to prevent agents from generating panic/log.Fatal/Must* code.
X-Lerian-Ref: 0x1
Agents were copying panic(err) patterns from code examples into business
logic. The examples are correct (panic IS acceptable in bootstrap/init),
but without explicit comments, agents treated them as general-purpose
patterns.
Added '// bootstrap-only: panic is acceptable in main/init; NEVER use
panic in business logic' comments to all panic() occurrences in:
- pm-team/skills/shared-patterns/code-example-standards.md
- dev-team/docs/standards/sre.md
- dev-team/docs/standards/golang/bootstrap.md
- dev-team/docs/standards/golang/core.md
- dev-team/docs/standards/golang/security.md
- platforms/opencode/standards/sre.md
Reported by: Garzão (#the-ring-feedbacks)
Refs: Ring agents generating panic() in service/handler code
Replace invented struct-field tracer (s.tracer.Start) with the actual
pattern from plugin-br-bank-transfer-jd: global OTel tracer via thin
wrapper package (telemetry.StartSpan).
- All s.tracer/h.tracer/r.tracer.Start → telemetry.StartSpan
- Add mandatory Telemetry Wrapper Package section with full tracer.go
- Update ASCII diagrams, instrumentation checklist, method template
- Update anti-patterns table to reference wrapper package
- Fix MIG-004 tracer criteria to match real pattern (global, not DI)
X-Lerian-Ref: 0x1
- Replace libUncommons alias with libCommons across golang.md and migrate-v4 skill
- Add explicit _test.go coverage to every MIG task acceptance criteria
- Add Task Generation Rule: files list MUST include test files
- Fix MIG-002 no-op line (libCommons alias stays, only import path changes)
X-Lerian-Ref: 0x1
- Re-add FORBIDDEN rule for direct Fiber response methods (chttp wrappers required)
- Fix bare tracer variable in pagination repository example (tracer → r.tracer)
- Standardize all error logging to clog.Err(err) instead of clog.Any("error", err)
- Remove cassert from MIG-003 acceptance criteria (de-para only, not new patterns)
- Add concrete OTel middleware setup in router example (chttp.NewTelemetryMiddleware)
X-Lerian-Ref: 0x1
Introduces a new parallel reviewer, `ring:dead-code-reviewer`, to
systematically identify orphaned code resulting from changes. This
enhances codebase hygiene by detecting unused helpers, types, and
tests, thereby reducing technical debt and improving maintainability.
The parallel code review system is expanded from 6 to 7 concurrent
reviewers. All related workflows, skills, commands, and documentation
have been updated to integrate this new capability.
- Add the `ring:dead-code-reviewer` agent, which analyzes code across
three concentric rings (target, direct dependents, ripple effect) to
find unreachable code.
- Update the `/ring:codereview` command and `ring:requesting-code-review`
skill to dispatch all 7 reviewers in a single parallel step.
- Adjust project documentation, including architecture diagrams, manuals,
and agent guides, to reflect the 7-reviewer workflow.
feat(default): add session-handoff skill to enable seamless context resume
refactor(arch): move complex orchestration from commands into skills
feat(pm-team): add orchestrator skills for pre-dev planning workflows
feat(dev-team): add skill to manage development cycle state (status/cancel)
feat(default): add skill to formalize git commit generation process
refactor(default): prune granular skills to focus core library on workflows
fix(pmo-team): align pmo-retrospective skill name with ring convention
docs: update all documentation to reflect major skill refactoring
Removes the hard-coded `model: "opus"` requirement from all agent
definitions, skill examples, and command documentation. The orchestrator
is now responsible for model selection, making the agent framework more
flexible and model-agnostic.
Key changes include:
- Removing the `model` parameter from all agent frontmatter and Task
tool examples.
- Deleting the "Model Requirement" hard gate and self-verification
sections from all agent prompts. This simplifies the prompts and
removes model-specific enforcement logic.
- Updating core documentation (ARCHITECTURE, MANUAL, README) to reflect
that model selection is an orchestrator concern.
- Removing redundant `changelog` and `last_updated` fields from agent
frontmatter, as this information is tracked by source control.
This architectural shift allows the orchestrator to dynamically choose
the best model for a given task based on complexity, cost, or speed,
without being constrained by the agent's definition. It simplifies
agent maintenance and prepares the system for easier integration of
future AI models.
