chore: release v0.2.6

[eggai-release-bot]

Release v0.2.6

Added

• baseUrl field added to aiStageConfig schema for openai-compatible providers, with OPENAI_BASE_URL / OPENAI_API_KEY env-var fallbacks resolved in getResolvedStageConfig.
• openai-compatible provider support for agentic review mode via @eggai/configurable-agent (Vercel AI SDK v5 agent loop). Any provider with a custom baseUrl can now run the full agentic security audit without SDK-specific adapters.
• Provider-dialect smoke spec: npm run test:smoke runs the 4 AI caller stages migrated in PR feat: native LLM structured responses via zod-driven complete #145 (file-reviewer, validation-resolver, dedup-resolver, root-cause-extract) against each real provider (anthropic, openai, bedrock, github) using a slice fixture as input. Validates that the structured-output dialect path returns a zod-validated response without throwing. Implemented as a Jest spec under tests/smoke/ with its own jest.smoke.config.ts — not picked up by default npm test (whose roots are limited to tests/unit/). Provider config comes from ConfigService + the existing PROVIDER_DEFAULTS table, not a duplicated table. Providers with missing credentials are describe.skip()-ed; providers with malformed credentials fail loudly via the provider class's own validateApiKey(). Input is a slice fixture under evals/datasets/inbox/smoke-sql-injection/, loosely following TDR 0002. Nightly + manual CI workflow at .github/workflows/provider-dialect-smoke.yml. Automates the unchecked manual smoke item from PR feat: native LLM structured responses via zod-driven complete #145's test plan; distinct from the deferred per-stage golden-evals item which validates output quality.
• unstructured dialect for LLM models without response_format: {type: "json_schema"} support (e.g. Llama, Qwen2.5, DeepSeek-V3, Phi, older o1-series). When isUnstructured() is true, the pipeline runs a full prose path: ProseFileReviewer → ProseValidationResolver → ProseDeduplicationResolver → session/prose-report.md. No JSON parsing, no structured schemas — the model writes free-form prose and subsequent stages refine it in-kind.
• Bundled litellm model capability snapshot (src/ai/providers/model-capabilities.json, 2101 chat models) for automatic supportsResponseSchema lookup. Unknown models default to unstructured (safe).
• scripts/update-model-capabilities.ts maintenance script to sync the snapshot from upstream litellm (npm run maintenance:update-model-capabilities [--write]).

---

Checklist:

• Version bump looks correct
• Changelog entries are accurate
• CI passes

After merging, the release will be automatically tagged and published to npm.

[github-actions]

QualOps Code Quality Analysis

Status: ✅ PASSED - No issues found

Summary

• Total Issues: 0
• Critical: 0 🔴
• High: 0 🟠
• Medium: 0 🟡
• Low: 0 🟢
• Files Analyzed: 2

No issues found in the analyzed code.

📊 Full Report

View detailed report

---

Powered by QualOps