Skip to content

OMK

v0.90.8 Feature

This release adds 3 notable features for engineering teams evaluating rollout.

Published 13d AI & Machine Learning
✓ No known CVEs patched
Read the diff → Tool health → What is this tool? →

✓ No known CVEs patched in this version

Topics

agent-orchestration agent-runtime ai-coding claude-code cli code-review
+14 more
codex coding-agent developer-tools devtools evidence local-first local-llm mcp model-context-protocol multi-agent opencode software-engineering typescript workflow-automation

Affected surfaces

auth rce_ssrf

Summary

AI summary

Updates New Features, docs/providers.md, and /../agent/README.md across a mixed release.

Changes in this release

Security High

Hardens guardrails evidence ledger with tamper‑evident hash chain, validates contract shape, returns deep copies, and escapes untrusted claim/command text in verification reports.

Hardens guardrails evidence ledger with tamper‑evident hash chain, validates contract shape, returns deep copies, and escapes untrusted claim/command text in verification reports.

Source: llm_adapter@2026-07-13

Confidence: high

Feature Low

Adds GPT-5.6 MoA virtual model combining concurrent Sol and Terra analysis.

Adds GPT-5.6 MoA virtual model combining concurrent Sol and Terra analysis.

Source: llm_adapter@2026-07-13

Confidence: high

Feature Low

Adds `partitionToolBatchWaves` to run ordered, contiguous safe tool-call waves instead of serializing entire batch on conflict or unknown call.

Adds `partitionToolBatchWaves` to run ordered, contiguous safe tool-call waves instead of serializing entire batch on conflict or unknown call.

Source: llm_adapter@2026-07-13

Confidence: high

Feature Low

Adds `compaction.model` setting allowing auto-compaction to use authenticated model like `zai/glm-5.2` without changing active session model.

Adds `compaction.model` setting allowing auto-compaction to use authenticated model like `zai/glm-5.2` without changing active session model.

Source: llm_adapter@2026-07-13

Confidence: high

Feature Low

Adds global-only `contextBudget.enabled` to activate context-budget v2 for every OMK session; per-process overrides remain via `OMK_CONTEXT_GOVERNOR=1/0`.

Adds global-only `contextBudget.enabled` to activate context-budget v2 for every OMK session; per-process overrides remain via `OMK_CONTEXT_GOVERNOR=1/0`.

Source: llm_adapter@2026-07-13

Confidence: high

Feature Low

Adds project-local Stagehand computer-use extension and `omk-computeruse` skill for bounded browser observation, approval‑gated actions, and redacted extraction.

Adds project-local Stagehand computer-use extension and `omk-computeruse` skill for bounded browser observation, approval‑gated actions, and redacted extraction.

Source: llm_adapter@2026-07-13

Confidence: high

Bugfix Medium

Fixes CLI help omission of accepted `max` and `ultra` thinking levels and resolves GPT-5.6 Codex `ultra` requests failing with invalid‑enum HTTP 400.

Fixes CLI help omission of accepted `max` and `ultra` thinking levels and resolves GPT-5.6 Codex `ultra` requests failing with invalid‑enum HTTP 400.

Source: llm_adapter@2026-07-13

Confidence: high

Bugfix Medium

Fixes loop write‑scope validation to collapse `.`/`..` path segments, preventing traversal like `src/../package.json` from bypassing scope checks.

Fixes loop write‑scope validation to collapse `.`/`..` path segments, preventing traversal like `src/../package.json` from bypassing scope checks.

Source: llm_adapter@2026-07-13

Confidence: high

Bugfix Medium

Fixes context‑budget v2 representation chooser to respect remaining tier budget, falling back to smaller representation instead of dropping whole item at ceiling.

Fixes context‑budget v2 representation chooser to respect remaining tier budget, falling back to smaller representation instead of dropping whole item at ceiling.

Source: llm_adapter@2026-07-13

Confidence: high

Bugfix Medium

Fixes footer `PKG` counter and control‑panel `ports:` label to correctly count accepted advisory/report‑only package ports, showing full connectivity (e.g., `PKG 16/16`).

Fixes footer `PKG` counter and control‑panel `ports:` label to correctly count accepted advisory/report‑only package ports, showing full connectivity (e.g., `PKG 16/16`).

Source: llm_adapter@2026-07-13

Confidence: high

Bugfix Low

Fixes Correctness Wall edit/write hooks to use fixtureless live OA only with non-empty run IDs, explicit MCP transport, and a bound MCP handler; unbound handlers fall back to preview‑only evaluation.

Fixes Correctness Wall edit/write hooks to use fixtureless live OA only with non-empty run IDs, explicit MCP transport, and a bound MCP handler; unbound handlers fall back to preview‑only evaluation.

Source: granite4.1:30b@2026-07-14-audit

Confidence: low

Full changelog

New Features

  • GPT-5.6 MoA virtual modelopenai-codex/gpt-5.6-moa combines bounded concurrent Sol and Terra analysis into one Sol-synthesized response. See provider setup.
  • Path-safe parallel tool-batch waves — independent tool calls run in ordered waves while conflicts and unknown operations remain sequential. See the agent runtime.
  • Context-budget v2 controls — enable the global budget with contextBudget.enabled and select an authenticated compaction model with compaction.model. See settings and compaction.
  • Evidence-gated Correctness Wall — fixtureless live OA uses a bound MCP handler only and otherwise remains preview-only. See the Correctness Wall example.
  • Project-local computer use — the Stagehand extension and omk-computeruse skill add bounded, approval-gated browser workflows. See the extension.

Added

  • Added the tool-free openai-codex/gpt-5.6-moa virtual model, which combines bounded, concurrent GPT-5.6 Sol and Terra analyses into one Sol-synthesized response for analysis and review tasks.
  • Added partitionToolBatchWaves to run ordered, contiguous safe tool-call waves instead of serializing an entire batch when one call conflicts or is unknown.
  • Added compaction.model, allowing auto-compaction and /compact to use an authenticated model such as zai/glm-5.2 without changing the active session model.
  • Added the project-local Stagehand computer-use extension and omk-computeruse skill for bounded browser observation, approval-gated actions, and redacted extraction.

Changed

  • Added global-only contextBudget.enabled to activate context-budget v2 for every OMK session; OMK_CONTEXT_GOVERNOR=1/0 remain per-process on/off overrides. Each enabled AgentSession reuses an in-memory plan and representation cache without disk or cross-session sharing.

Fixed

  • Fixed the CLI help omitting the accepted max and ultra thinking levels, and fixed GPT-5.6 Codex ultra requests failing with an invalid-enum HTTP 400.
  • Fixed loop write-scope validation to collapse ./.. path segments, so traversal like src/../package.json can no longer slip past allowedWriteScopes, deniedWriteScopes, or changed-file checks.
  • Fixed the context-budget v2 representation chooser to respect the remaining tier budget: it now falls back to a smaller representation (e.g. summary) instead of over-selecting full text and dropping the whole item at the tier ceiling.
  • Hardened the guardrails evidence ledger: replay events now carry a prevHash/eventHash tamper-evident chain verified on load (edited, deleted, or reordered ledger lines fail closed), TaskContractBuilder.fromJSON validates the contract shape instead of blindly casting, build()/getEvents()/getLedger() return deep copies, and verification-report table cells escape pipes/newlines from untrusted claim or command text.
  • Fixed the footer PKG counter and control-panel ports: label dropping accepted advisory/report-only package ports from every bucket (PKG 12/16 R0 B0 for 16 accepted candidates); the intake summary now counts them via acceptedAdvisory, so all accepted ports read as connected (PKG 16/16).
  • Fixed Correctness Wall edit/write hooks to use fixtureless live OA only with non-empty run IDs, explicit MCP transport, and a bound MCP handler; unbound handlers fall back to preview-only evaluation.

Weekly OSS security release digest.

The CVE patches and breaking changes that affected production tools this week. One email, every Sunday.

No spam, unsubscribe anytime.

Share this release

Track OMK

Get notified when new releases ship.

Sign up free

About OMK

Provider-neutral CLI control plane for coding agents: routes runtimes, scopes MCP, runs DAG workers, and verifies evidence before completion. MIT.

All releases →

Related context

Related tools

Beta — feedback welcome: [email protected]