Skip to content

Releases: Zhivex/zhivex-harness

v1.2.0-rc.8

v1.2.0-rc.8 Pre-release
Pre-release

Choose a tag to compare

@mortiz-dev mortiz-dev released this 26 Sep 15:51
d771d4a

Programming assistant continuity and OCI delivery

RC8 includes the fixes accumulated since the published RC5:

  • Meta stateless encrypted reasoning replay, avoiding dependence on previous-response availability.
  • Qwen reasoning deduplication and adaptive context compaction that preserves the newest indivisible interaction.
  • Clearer coding-tool guidance and readable terminal Markdown tables.
  • OCI delivery checks before completion, with bounded read-only recovery and normal import approvals.
  • Parent completion checks for pending edits in delegated child and descendant snapshots.
  • Specific sanitized diagnostics for incomplete delivery and live approval/resume failures.

This is a prerelease on npm next; the stable latest channel remains unchanged.

bun add --global @zhivex-ai/harness@1.2.0-rc.8
zhx --version

Release evidence is bound to the exact tarball, annotated tag, source commit and protected workflow. RC6 and RC7 were not published; their immutable tags and failure evidence remain historical records.

Verification

  • Source: d771d4a6dc2d8aeb5fb260c4eb51ac761f560f9d.
  • Exact-artifact validation and live certification passed.
  • Representative matrix: 42/42 cases passed across Meta contributor, Qwen Flash and OpenAI Luna.
  • npm next, exact tarball SHA-512 and SLSA provenance independently verified.
  • The initial postpublication check timed out awaiting npm attestation propagation. A retry of only the publish/verify job detected identical registry bytes, skipped publication and verified successfully.
  • Workflow: https://github.com/Zhivex/zhivex-harness/actions/runs/36251780208

1.2.0-rc.5

1.2.0-rc.5 Pre-release
Pre-release

Choose a tag to compare

@mortiz-dev mortiz-dev released this 26 Sep 11:36
9553949

Prerelease candidate for the next channel.

Install: bun add @zhivex-ai/harness@1.2.0-rc.5

  • Adapt the assistant loop to explanation, investigation, design and implementation without imposing a repair ceremony on read-only work. Remove the strict/repair runtime profile choice; callers requiring verified completion use requireVerifiedDelivery or --require-verified-delivery directly. This is an intentional pre-promotion API/CLI break; saved connection profiles and subagent roles remain independent.
  • Unify bounded tool-error, invalid-argument and unknown-tool recovery across CLI, SDK and delegated runs, independently of the repair controller. Explicit fail-fast tool options remain available.
  • Return malformed application-owned delegation arguments through bounded SDK validation feedback without launching a child. Preserve hidden-tool rejection and exact task-ID authorization.
  • Preserve conversational decisions and bounded, redacted local error observations through repeated deterministic and model-assisted compaction. Context is working memory; approvals and verification remain tool-owned.
  • Prioritize recent user corrections over the historical objective and assistant recollections during compaction; retain more distinct user steering within the same summary budget.
  • Use the remaining repair output budget instead of an implicit 2,048-token per-response cap. Preserve explicit caps and provider transport compatibility. Finish after relevant checks and change review without requiring duplicate audits.
  • Refresh discovered directory guidance after normal edits instead of aborting the conversation; retain path, link, encoding and size protections.
  • Return the matching persisted context/accounting checkpoint in all runs, and detect repeated tool cycles even when the model changes its progress narration.
  • Invalidate failed or superseded model edit references and explain how to recover them before another approval. Keep partially failed changes pending verification and preserve closure capacity after rejected plans or out-of-scope reads.
  • Remove lifetime cutoffs for operator requests and discovered instruction counts; paginate task history and preserve repeated latest requests such as A → B → A. New user requests reset loop history while approval resumes retain it.
  • Share a finite 50-step default across API and chat, and accept explicit larger step/tool-call limits without the former arbitrary ceilings.
  • Preserve the newest correction before the historical objective under tight summary budgets, including corrections after a previous summary emptied the objective. Authenticate builtin output before semantic export; restored summaries retain user objectives and steering without exporting unproven tool diagnostics or excerpts.
  • RC5 passed exact-artifact certification and was published to npm under next. RC4 remains unpublished, with its tag and failure evidence preserved.

Source: 9553949282e4ac70cda9f4103a454ac42cab1131.

Exact-artifact certification and publication workflow.

Verification: 1,358 local tests passed (one platform skip); protected exact-artifact, live provider, and representative gates passed. The representative matrix passed 42/42 cases: Meta muse-spark-1.3, Qwen qwen3.8-flash, and OpenAI gpt-6-luna. Registry integrity, the next tag, and SLSA workflow provenance were independently verified. Gemini remains outside this certification cohort.

Zhivex Harness 1.0.0

Choose a tag to compare

@mortiz-dev mortiz-dev released this 20 Sep 14:35
69c8a4b

Zhivex Harness 1.0.0 is published to npm as @zhivex-ai/harness@1.0.0 on latest.

The stable runtime and dependencies match the reviewed RC14 baseline. Stable API/CLI/schema contracts and Node/macOS/Linux support follow the repository's stability and support policies. Meta, Qwen and OpenAI each passed all 14 representative cases against the exact stable artifact; Gemini remains provisional.

The protected release passed deterministic and installed-artifact validation, OCI and live certification, all 42 representative cases, and registry integrity/SLSA verification. Publication used the original validated tarball at commit 69c8a4b0b3e62c3f492257a792670ce6ce8f4298.

Attempt 1 failed one Qwen representative case and did not publish. The explicitly authorized second attempt reran the complete matrix and passed; the original failure remains documented. This does not erase the separate exploratory SWE-bench result of Harness 0/5 versus control 1/5 or establish broad coding accuracy.

The final RC14 security assessment was performed by Codex under an explicit user-delegated AI exception, not an independent human audit. Four known findings were mitigated; documented host, service, redaction and verifier risks remain.

Existing paused approvals require the original artifact and policy context. Consult the stability, migration and rollback policies before changing versions.

Release workflow: https://github.com/Zhivex/zhivex-harness/actions/runs/35514103845

SHA-512 integrity: sha512-wIotQy/Az7UsfSfQyQusQZ5PJFBuOTWJoylLVRMsYnDKtF1MNuOVALgNFSAmrfdzEOUrUhU4EmWBpf8RC8UVyA==

1.0.0-rc.13

1.0.0-rc.13 Pre-release
Pre-release

Choose a tag to compare

@mortiz-dev mortiz-dev released this 08 Sep 18:58
d6e310c

RC.13 adds secure provider/model profile onboarding and the interactive daily workflow: bounded multiline input and history, context and attachments, terminal-safe presentation, and durable interruption/approval recovery.

The protected release passed deterministic and installed-artifact checks, OCI boundaries, live certification, and 14/14 representative cases for each of Meta, Qwen, and OpenAI. Published to npm next; exact integrity and SLSA provenance were independently verified.

This is a prerelease. The independent human security review remains required before 1.0.0 GA.

Zhivex Harness 0.11.1

Choose a tag to compare

@mortiz-dev mortiz-dev released this 23 Aug 15:26
d7e3e87

What's Changed

  • Harden harness workflows for release readiness and live certification by @mortiz-dev in #16
  • chore(deps): bump actions/upload-artifact from 4.6.2 to 7.0.1 by @dependabot[bot] in #21
  • chore(deps): bump actions/download-artifact from 4.3.0 to 8.0.1 by @dependabot[bot] in #20
  • chore(deps-dev): bump @types/node from 24.13.3 to 26.2.0 by @dependabot[bot] in #19
  • chore(deps): bump actions/setup-node from 6.2.0 to 7.0.0 by @dependabot[bot] in #17
  • Harden harness security and release readiness workflows by @mortiz-dev in #22
  • Release @zhivex-ai/harness 0.11.1 by @mortiz-dev in #23
  • Harden release evidence validation by @mortiz-dev in #24

New Contributors

Full Changelog: v0.11.0...v0.11.1

Zhivex Harness 0.11.0

Choose a tag to compare

@mortiz-dev mortiz-dev released this 22 Aug 10:34
652a299

What's Changed

Full Changelog: v0.10.0...v0.11.0