Changelog

All notable changes to this project will be documented in this file.

The format is based on Keep a Changelog.

Unreleased

Added

• Add scaled signal_to_metrics correctness guidance and regression coverage for producer identity, backend resource mapping, temporality, and dashboard aggregation

Changed

• Replace upstream digest issue updates to overwrite issue body snapshots (instead of appending) while preserving idempotent no-op reruns on exact body matches
• Update GenAI guidance for the separate conventions repository, provider-versus-agent identity, histogram token usage, current content attributes, and execute-tool span naming
• Refresh AI-agent telemetry guidance from current upstream documentation: Claude Code beta traces, Gemini prompt/traces controls, Copilot namespace migration, and Codex mode-specific verification
• Add current OpenTelemetry blog playbooks for Go compile-time instrumentation v1, GenAI observability, and OTel blueprints
• Update instrumentation guidance from deprecated gen_ai.system to gen_ai.provider.name and document stable Go compile-time instrumentation as an alternative to eBPF and manual SDK work
• Migrate Collector examples and evaluations to canonical component IDs and raise the compatibility floor to v0.153.0
• Document sparse exporter-helper failure metrics and expand upstream watcher coverage for GenAI conventions and exporter helpers
• Bump aligned Tessl skill and tile metadata from 0.4.0 to 0.5.1

[1.3.0] - 2026-05-08

Added

• references/anti-patterns.md: full annotated catalog of 17 anti-patterns organized by category (pipeline, Kubernetes, metrics, AI agents, OTTL)
• Progressive Disclosure trigger for anti-patterns.md in SKILL.md

Changed

• Teach the skill to audit existing collector/Helm configs for cross-field contradictions such as memory limiter sizing, sticky routing for scaled tail sampling, hostPort misuse, durability gaps, rollout inconsistency, OTTL type drift, unsafe RWX/NFS-backed queues, and risky temporality conversions
• Trim SKILL.md inline anti-patterns from 17 to 6 most critical; remaining 11 moved to references/anti-patterns.md
• Clarify hook support and governance in AI agent observability reference
• Document limitations, roadmap, and contribution guidelines in README, CONTRIBUTING, and this changelog
• Make bundled references and evaluation docs review-visible via a Tessl docs entrypoint
• Move fast-changing compatibility details out of SKILL.md and into references/compatibility.md
• Add practical upstream watch guidance for bbolt security advisory tracking and event-to-logs proposal maturity

1.2.0 - 2026-03-20

Added

• AI coding agent observability reference (references/ai-agents.md) with support matrix for Claude Code, Gemini CLI, GitHub Copilot, Codex CLI, and others
• Recommend opentelemetry-hooks for agents without native telemetry
• Production playbooks routing reference from OpenTelemetry.io blog series (references/playbooks.md)
• 5 new test scenarios for AI agent observability

Fixed

• Homepage references and installation URLs

1.1.0 - 2026-03-11

Added

• Connectors reference (references/connectors.md): span_metrics, service_graph, routing, and failover with stickiness requirements
• Kafka resiliency patterns, semconv v1.30+ guidance, config management, scaling guidance (Collector v0.147.0)
• Upstream maintenance digest workflow (weekly automated scan of OTel issues, releases, and blog posts)
• Cursor Marketplace plugin manifests
• Markdownlint configuration for documentation style
• Filesystem compatibility warnings for file_storage and bbolt
• Document semconv handling for intentional HTTP cancellations
• Load-balancing routing key requirements and Go SDK 1.40.0 breaking changes

Changed

• README trimmed: removed Terraform comparison table, fixed installation URLs
• Applied upstream maintenance guidance from 2026-02-24 digest
• bbolt v1.4.3 on_rebound crash mitigation documented

1.0.0 - 2026-02-03

Added

• Initial cognitive router (SKILL.md) with System 2 Thinking framework
• 11 progressive disclosure triggers mapping to reference documents
• Reference documents: architecture, collector, instrumentation, sampling, security, monitoring, OTTL, platforms
• Anti-pattern prevention with explicit language (MUST/NEVER/ALWAYS)
• TDD testing framework with rationalization table
• Claude marketplace plugin (.claude-plugin/marketplace.json)
• CI validation workflow (frontmatter, links, markdown linting)
• Auto-delete merged branches workflow
• Contrib component references and example configs
• Per-language SDK links and instrumentation packaging guidance