@jyleaves/dsh-win-bash-fix
Windows Git Bash fix for DeepSeek Harness (dsh): MSYS sandbox conflict + terminal inspector, plus the minimal-win collective-thinking preset (we/let's reasoning chain).
92 results
Windows Git Bash fix for DeepSeek Harness (dsh): MSYS sandbox conflict + terminal inspector, plus the minimal-win collective-thinking preset (we/let's reasoning chain).
nono (Landlock/Seatbelt) backend for the DeepSeek Harness sandbox seam with a vendored fail-closed executor
Agent-decided approvals for DeepSeek Harness: a workspace-write base permission mode where an independent approval subagent judges every sandbox escalation (risky operations are rejected), with a configurable approval model and a per-session audit trail in the conversation window's 审批 tab.
A middle permission tier for DeepSeek Harness between workspace-write and danger-full-access: auto-approves harmless commands and operations targeting configured trusted areas, beyond the current workspace.
Real-time attention alerts for DeepSeek Harness: a global banner plus flashing workspace entries when the agent is blocked (approval request / sandbox denial), with an audible chime and OS-level notifications for new blockers.
DeepSeek Harness plugin (temporary, pre-upstream-fix): automatically injects a CPython tempfile shim (PYTHONPATH -> sitecustomize) into every confined shell command on the Windows sandbox, so python/pytest tempfile use works with zero extra tools, zero model-context overhead, and zero escalation.
Retrieved ≠ injected: CBDC-gated memory for DeepSeek Harness — decides how memory is USED (use/verify/ignore decisions, feedback learning, full audit); local SQLite + FTS5, bounded, no extra model call
Seatbelt (libsandbox) sandbox provider for the DeepSeek Harness sandbox seam on macOS: a cordis plugin that replaces @deepseek-ai/dsh-sandbox-local's seatbelt rung via a prebuilt seatbelt-run launcher — no dsh source changes, no PR required
Workspace Only switch for DeepSeek Harness: confine file access of every sandbox mode to the session workspace
NeevSandbox execution providers for DeepSeek Harness
Standalone DeepSeek Harness access-mode plugin with Default, No Edit, and Auto modes
DSH plugin: strip GPT-family tool-call sandbox_permissions that are not strictly wider than the current session.
V0 capability-disclosure scanner for DeepSeek Harness (DSH) cordis plugins: reports which sensitive services (credentials/shell/approval/sandbox/fs/tools...) a plugin declares or references before you trust it.
Tiered auto-review for DeepSeek Harness (DSH): static-rule safety net + LLM reviewer + human fallback. Auto-approve safe actions, auto-deny irreversible ones, ask a human for the rest. PURE VIBE CODING - not audited, use at your own risk.
DSH microsandbox bundle: Linux microVM shell executor and ctx.sandbox provider (msb run) with workdir/env preservation, fail-closed probing, and shell-free argv construction
Runtime security guard for DeepSeek Harness (DSH) — blocks command injection, SSRF, credential exfiltration, and destructive operations at runtime.
Better config hub for DeepSeek Harness: one page showing sandbox mode, default model, workspaces, settings namespaces and copy-ready composition templates.
One-click control that closes a session's persistent bash terminal (dual-face: /exit-bash command + session-header button) so the sandbox mode can be switched
Fail-closed DSH compatibility guard for redundant GPT/Codex sandbox escalation arguments
Strip sandbox_permissions / justification from model tool-call arguments so sandbox escalation is never triggered when the session already has sufficient permission.
DSH profile plugin: unconfined (GPU-capable) session sandbox plus per-operation user approval for writes outside the workspace or to protected files.
AgentENV execution-world bundle for DeepSeek Harness
Approval timeout for DeepSeek Harness: when a sandbox/escalation approval is not answered within a configurable period, the request is automatically rejected instead of waiting forever.
Self-protection guardrails for the DeepSeek Harness: protected-path interception, backup-before-destroy, CLI self-recovery commands, and strict sandbox defaulting