dsh-sandbox-permissions-not-strictly-wider-justification-empty-fix
修复 sandbox_permissions 和 justification 字段在 pwsh/bash/fs/dsh-sandbox 中的 no-op 升级报错。当会话已是 danger-full-access 模式时,模型携带空 justification 或重申同一模式被拒的问题。
76 results
修复 sandbox_permissions 和 justification 字段在 pwsh/bash/fs/dsh-sandbox 中的 no-op 升级报错。当会话已是 danger-full-access 模式时,模型携带空 justification 或重申同一模式被拒的问题。
Scan project lockfiles (npm/pnpm/yarn/pip/go/cargo/maven/gradle/...) against the free OSV API and report confirmed dependency vulnerabilities with fix versions and remediation commands.
Configurable auto-approval for DeepSeek Harness: intercepts approval/request and answers allowed-once/rejected by policy (allow-all / allowlist / deny-always), with a full audit log — so the human is only asked when the policy says so.
A DeepSeek Harness Web plugin for inspecting redacted LLM HTTP request and response traces.
Durable agent-to-agent mailbox for DeepSeek Harness: any MCP client or DSH session can address any other. Threads, receipts, search, broadcast, edit/withdraw, reactions. Local-only, zero dependencies, secrets redacted on ingest.
DeepSeek Harness plugin: run a live security health check on the deployment (credential file permissions, process-env secrets, permission/approval mode, telemetry, anonymous id, LAN exposure, sandbox backend) and get an actionable report mapping each finding to upstream discussions.
Tolerate redundant sandbox_permissions requests: a tool call asking to escalate to a mode that is not strictly wider than the session's standing mode runs as-is at the standing mode instead of failing with "not strictly wider". Genuine escalations still route through the approval flow.
Approval-free writes into whitelisted directories: wraps ctx.fs.checkedTarget for write/edit, auto-approves bash escalations that touch whitelisted paths, with a Settings GUI. 白名单目录写入免审批(write/edit + bash),设置页可视化调整。
Workspace tools for DeepSeek Harness: per-workspace default Agent preset and permission preset for new sessions, using only official extension points (no DSH patches required).
DSH compatibility guard: repairs compaction limits, fail-safe sandbox escalation, and missing tool descriptions for third-party models; capability lookups are cached and the deprecated prompt-injection path is off by default.
Permission matrix for DeepSeek Harness: 3 sandbox modes x 4 approval strategies = 9 selectable presets, with a global default, a separate LLM-robot default, four-tier risk policies, password-gated approval for high-risk operations, audit log and git checkpoint.
Audit a DeepSeek Harness session against the 14 consciousness indicator properties of Butlin, Long, Elmoznino, Bengio et al. (2023) and report an L1-L5 level
Three-mode visible-session advisor plugin for DeepSeek Harness with configurable tool permissions, local-subagent coverage, and task-tree budgets.
DSH Metafolder Plugin — visual meta-folders for the DeepSeek Harness sidebar workspaces: nest real workspace folders under named, collapsible groups, by drag and drop or the row menu. Paths, permissions, sessions and workspace ordering are untouched.
Self-healing DeepSeek-Harness plugin: keeps confined pwsh/bash console windows invisible (SW_HIDE) on Windows — no black console flash on every sandboxed command.
Normalize invalid non-widening sandbox escalation arguments before DSH tool dispatch.
A fourth permission option for DeepSeek Harness sessions: selected-workspace-write, where the user opens the workspace directory tree and toggles which directories the agent may write to.
Cross-provider second-opinion advisor for dsh. v2 (simple-extensible-configurable-stable): DEFAULT ROUTE = official DeepSeek (deepseek-official/deepseek-v4-flash); model param optional everywhere; one-page visual settings panel; honest audit trail. Scout mode = official + live web_search. OpenAI-compatible endpoint + consult tool + DSH_MODEL rerouting.
Automatic review of sandbox escalation requests for DeepSeek Harness, plus an /approve slash command for manually approving auto-denied actions (Codex Guardian-style)
DSH++ — a local-first control plane for DeepSeek Harness. Plugin security auditing and workflow regression testing for dsh, in a CLI plus a loopback web console.
Quality-of-life plugins for DeepSeek Harness: workspace rewind, desktop notifications, auto-format, live git context, secret redaction, and a done-means-green verify gate.
Audit a DSH profile for capabilities nothing consumes, and record completion claims with the evidence that supports them.
AI-reviewed auto-approval for DeepSeek Harness: judges each pending approval ask with one auxiliary model call before any human answerer runs, auto-grants the safe ones, and escalates everything else with the review analysis attached
Put DeepSeek Harness in your pocket: approve tool calls and answer ask_user_question prompts from your phone over Feishu, with the desktop always first in line.