为下一个工作流
找到对的 DSH 插件

Frame-level scan diagnostics for session files (torn/corrupt/empty detection).

Self-evolution: the agent hot-mounts/removes persistent plugins on itself mid-session.

Telemetry backend exporting turns, model steps, and tool calls to yiTrace.

Exposes bounded Prometheus metrics for DSH sessions, agents, LLM calls, tools, approvals, subagents, and jobs, with a Grafana dashboard and a loopback-only endpoint by default.

Redacts supported secret patterns from the `session-telemetry/record` export copy before configured telemetry backends receive it.

Writes local JSONL summaries of per-turn tool counts and coarse verification signals without storing prompts, tool arguments, or result text.

Appends local hash-chained JSONL receipts for tool results and session events without storing prompts, tool arguments, result text, or raw session IDs.

Support for the microsandbox backend.

Microsoft cross-platform sandbox support.

Support for the nono sandbox backend.

Agent-tree token budget management.

The model switch for DSH: point it at any OpenAI-compatible endpoint, with curated free/cheap DeepSeek provider presets and automatic fallback when a free tier rate-limits you.

Fine grained per tool permission rules for DSH at the tools/pre-execute gate, deny and ask lists in Claude Code rule syntax (Bash(rm -rf:*), Read(*secrets*), mcp__server__tool), works standalone without migrating.

Manual approval mode ("Manual Mode" / "Ask Mode").

Turn-scoped “Allow for this task” approvals: automatically allow matching `danger-full-access` escalations only for the current task, then expire.

Plugin template repo (based on the official turtle-ui repo).

A TPS metrics plugin.

Per-process tool-call statistics: a `tool_stats` tool reporting per-tool call counts, error counts, and average durations.

Auto-log failed tool calls across native tools, PTC run_code, and inline invocations: dedup and count root causes into a skill so repeated mistakes fade.

Zero-touch runtime telemetry for DSH: every session auto-writes task, model tier, tools, failures, duration, status into a human-readable journal with a stats section (pure observer — no tools, no LLM calls, no injection).

Evaluation harness for DSH plugins: YAML cases drive real headless agent runs, assert on tool calls, args, results and token usage, with a baseline gate for CI regression.

Community distribution: TUI, desktop, and Web UI as one bundle with layered installation.

Select browser elements directly during Vibe Coding and send structured visual feedback to the DeepSeek Harness Agent.

Reusable Markdown prompt profiles for DeepSeek Harness with per-turn model selection, argument substitution, and state restoration.