v0.3.0 — Prime Lab Integration

Highlights

• Prime Lab Integration (WP2.5): Hosted training and evaluation scaffolding with prime lab / prime env CLI support

  • TOML configs for hosted RL and eval (configs/rl/, configs/eval/, configs/endpoints.toml)
  • Makefile targets: lab-check, lab-run-e1/e2, lab-eval-e1/e2, env-eval-e1/e2
  • Compatibility gate script (scripts/prime_lab_check.py) with Rich output parsing fix
  • Metadata normalizer for hosted → SV-Bench schema (scripts/normalize_hosted_eval.py)

• WP2 Baselines & Public Mini Sets (#61): Baseline evaluation framework on public mini datasets

• Metrics & Reporting (#56, #58): Metrics contracts, E1/E2 report generators, summary.json auto-write

• Research: SOAR investigation (#63), MaxRL investigation for future experiments

Breaking Changes

None. All new functionality is additive.

Dependency Updates

• prime CLI: requires ≥0.5.37 for lab/env subcommands
• prime-rl v0.4.0 available as optional dependency (uv sync --extra lab)
• security-verifiers-utils bumped to 0.3.0

Full Changelog

v0.2.3...v0.3.0

v0.2.3

What's Changed

Fixes

• Handle markdown code blocks in model JSON outputs (models like qwen, llama wrap JSON in json...)
• Fix E1/E2 report scripts to find OpenRouter nested model directories

Improvements

• Add agent skills for common repo workflows (sv-eval, sv-data, sv-deploy, sv-hf, sv-dev)
• Update skill documentation with troubleshooting and workflows

Deployments

• security-verifiers-utils v0.2.3 → PyPI
• sv-env-network-logs v0.2.11 → Hub
• sv-env-config-verification v0.2.12 → Hub
• sv-env-code-vulnerability v0.2.5 → Hub

v0.2.2

What's Changed

• Prepare E3 vulnerability environment for production by @intertwine in #34
• Plan E4 phishing detection production deployment by @intertwine in #35
• Plan E5 production environment setup by @intertwine in #36
• Plan E6 Red-Team Defense production environment by @intertwine in #37
• Fix: robust response parsing, JSON patching, Weave init side-effects, E2 packaging, reward & tempfile fixes by @intertwine in #52
• chore: bump verifiers to >=0.1.9, sv_shared to 0.2.2 by @intertwine in #53
• chore: bump versions to v0.2.2+, require security-verifiers-utils>=0.2.2 by @intertwine in #54

Full Changelog: v0.2.1...v0.2.2

v0.2.1

Full Changelog: v0.2.0...v0.2.1

v0.2.0

Full Changelog: v0.1.5...v0.2.0

v0.1.5

What's Changed

• docs: add Prime Intellect Environments Hub compatibility analysis by @intertwine in #32
• feat: implement Prime Intellect Environments Hub compatibility by @intertwine in #33

Full Changelog: v0.1.4...v0.1.5

v0.1.4

Full Changelog: v0.1.3...v0.1.4

v0.1.3

Full Changelog: v0.1.2...v0.1.3

v0.1.2

Full Changelog: v0.1.1...v0.1.2

v0.1.1

Full Changelog: v0.1.0...v0.1.1