dsh-yolo-mode
LLM-powered auto-approval and policy plugin for DeepSeek Harness sandbox escalations with fail-closed safety.
Install
$ dsh plugin --profile web add dsh-yolo-modePlugin Overview & Capabilities
AI-assisted organization based on the public repository snapshot. The content must be grounded in source evidence and does not replace compatibility or security verification.
Key Capabilities
- Intercepts sandbox escalation requests and delegates decisions to an LLM judge based on configurable presets
- Provides fine-grained per-tool and per-target permission levels (allow, judge, delegate, deny)
- Ensures fail-closed security by defaulting timeouts, invalid outputs, or errors to human delegation or denial
- Offers web UI components including an input status chip, decision statistics popup, and runtime settings page
- Maintains local JSONL audit logs recording tool names, target modes, rationale, decisions, and error codes
Useful For
- Automating repetitive approval prompts during autonomous coding sessions while maintaining granular control
- Enforcing strict or balanced sandbox privilege escalation policies for multi-tool agent workflows
- Auditing and inspecting LLM agent permission escalation requests via structured JSONL logs and web UI analytics
Who It Fits
- Developers seeking hands-free DeepSeek Harness agent execution without completely bypassing sandbox security
- Teams requiring auditable, tiered sandbox escalation controls and safety fail-safes for autonomous agents
Documented Limitations
- High-reasoning LLMs require larger judge token budgets (judge.maxTokens) to prevent premature truncation and BAD_OUTPUT fallback errors
- Overlaps functionally with DeepSeek Harness experimental native auto-review (@deepseek-ai/dsh-experimental-auto-review)
- Author notes the plugin may be deprecated in the future in favor of official upstream auto-review capabilities
DSH Compatibility
Version-specific runtime evidence collected by DSH Plugin. A missing result means we have not tested that combination yet.
Security Signals
Objective signals discovered from package metadata and source inspection. These are not a guarantee that a plugin is safe.
package.json declares 0 runtime, 6 development, 9 peer, and 0 optional dependencies.
package.json declares DSH bundle metadata.
GitHub reports the repository license as MIT.
A root package.json was captured and can be inspected by the registry.
Public GitHub source metadata is available for this registry snapshot.
Source & Registry Notes
Public provenance, Registry classification, and the latest source check for this entry, kept separate from runtime verification.
- Source repository
- SeverusZh/dsh-yolo-mode
- Registry source
- GitHub · dsh-plugin topic
- Registry classification
- Plugin
- Source checked
- a83f46a · 2026-10-22
This project is independently indexed from public source information. DSH Plugin is not affiliated with DeepSeek or the plugin author. Always check the author repository before installation.
Repository Activity
- GitHub stars
- 3GitHub stars
- Forks
- 0
- Open issues
- 0
- Last commit
- 2026-10-19
- Last release
- 2026-10-19
Related DSH Plugins
Ranked by overlapping capabilities, use cases, plugin type, categories, and DSH profile.
Self-hosted AI agent runtime and MCP bridge providing sandboxed execution, memory, and audit logs.
View plugin →Native SQLite-backed task management and project board bundle with Web UI and Agent workflow tools for DeepSeek Harness.
View plugin →Persistent multi-model agent teams with dynamic DAG orchestration, run monitoring, and token insights for DeepSeek Harness.
View plugin →Free frontier LLM provider plugin for DeepSeek Harness with upstream probing and local OpenAI-compatible gateway.
View plugin →