Skip to content

V-Agent 1.1.0

Choose a tag to compare

@IsValorum IsValorum released this 14 Nov 22:03
· 11 commits to main since this release

V-Agent 1.1.0

Release focused on cleaner tool-execution UI, automatic minor updates without manual intervention, optional local engine setup on fresh installs, hardened remote API compatibility with per-connection UI controls, and clearer MCP parameter diagnostics.

Interface & Tool Trace

  • Collapsible tool trace (agent/ui/app/chat.js, panels.css): Tool executions inside assistant responses are grouped under <details class="tool-trace-group"> (collapsed by default) with a compact summary bar showing running step / error counts and individually expandable <details class="tool"> entries for inspecting arguments and outputs on demand.
  • Complete 5-language translations (agent/ui/locales.json): Added full translations (es, en, pt, fr, de) across 1,478 phrases including the onboarding tour, collapsible tool trace, optional engine step, and every new API connection/header control, placeholder, and validation message.

Automatic Minor Updates & Optional Local Engines

  • Hands-free minor updates (agent/desktop_runtime.py, agent/components.py, agent/ui/app/core-connection.js): Startup component checks run shortly after launch (1.5 s) with progressive backoff retries (20 s–300 s), fetch up to 100 releases per page with strict signed-manifest checks, hot-reload UI patches into the active window, and fire v-agent-components-update events automatically without requiring manual update clicks.
  • Optional local engine onboarding (agent/ui/app/tour.js, shell.js, engine_updates.py, engine_manager.py): First-run onboarding step 2 allows choosing llama.cpp, Strata, or Strata Rust, switching directly to an external API, or skipping engine installation cleanly; update banners are suppressed for engines that are not installed, and missing-engine errors distinguish Strata Rust from llama.cpp.

Hardened Remote API Protocols & UI Controls

  • Protocol & header hardening (agent/remote_protocols.py, agent/engine_manager.py, agent/agent.py, agent/ui/app/engine-controls.js):
    • Sends global User-Agent: V-Agent/1.1.0 by default (with per-connection override presets/custom value), strips X-Stainless-* SDK headers, and rejects CRLF/non-Latin-1 header values.
    • Supports provider presets and auth modes (auto, openai, anthropic x-api-key + anthropic-version + anthropic-beta, azure api-key + api-version + classic /openai/deployments/<dep> 404 fallback, opencode per-conversation persistent UUID v4 x-opencode-session across turns/restarts/compaction, and openrouter HTTP-Referer / X-Title / delta.reasoning / SSE comments).
    • Enforces SSE anti-buffering headers (Accept: text/event-stream, Cache-Control: no-cache, Accept-Encoding: identity) on streams with transparent gzip/deflate fallback decompression.
    • Exposes provider preset, auth mode, User-Agent, session header mode, Anthropic beta/version, Azure API version, HTTP-Referer, X-Title, and custom headers cleanly per connection in the UI menu.

MCP Diagnostics

  • pc MCP parameter errors & recovery (agent/pc_mcp.py, tests/smoke/mcp_recovery.py): Reports missing required KeyError parameter names clearly to the model so calls can recover cleanly on retry.

Installer & Verification

  • Clean-worktree NSIS zlib installer (V-Agent_1.1.0_x64-setup.exe, 260,348,539 bytes, SHA-256 ae455714d889e86e858e5f2c96078b1e686ace00d55bb7ccd8fc7323ec84961a) verified free of personal data or development artifacts, with Ed25519-signed v-agent-update.json and v-agent-update.json.sig.
  • Tested: Verified scripts/build-ui.py (1,478 translated phrases), static_analysis.py (no stray control chars across 271 text files; ruff/bandit/ESLint skipped as not installed), all 8 Node .cjs UI test suites (including test_ui_i18n.cjs and test_v2_ui_features.cjs after the final UI translation commit), api_edge_cases.py (19 checks, 0 failed), xss_model_card.py, language_checks.py (45 checks, 0 failed), and Clean release verification passed (clean_release_build.py). On Python unittest (308 tests), 307 passed and 1 test (test_pc_hf_web_admin_computer_tools_end_to_end) hit a local port 19380 collision because an existing desktop instance of V-Agent was running on the build host.