zhang8019/dsh-permission-matrix ↗★ 0
dsh-permission-matrix
Permission matrix for DeepSeek Harness: 3 sandbox modes x 4 approval strategies = 9 selectable presets, with a global default, a separate LLM-robot default, three-tier risk policies, audit log and git checkpoint.
安装
$
npx -p @deepseek-ai/dsh dsh plugin --profile web add github:zhang8019/dsh-permission-matrix说明文档
阅读完整 README ↗Configuration
Settings → Permission Matrix, or the permission-matrix row of /cordis.patch.yml:
| Key | Default | Meaning |
|---|---|---|
enabled | true | master switch |
takeover | see above | preset id → tier (auto-allow / classify) |
midRiskPolicy | deny | deny / allow / ask |
llmJudge | true | enable the LLM judge |
judgeProvider / judgeModel | empty | empty = follow the current session model |
judgeStages | both | both / fast / thinking |
autoAllowHardGuard | true | hard-deny protection in the auto-approve tier |
robotDefaultPreset | fa-auto | default preset for robot sessions |
robotWorkspaces | [] | robot session workspaces (absolute paths, subdirs included) |
gitSnapshot | true | git checkpoint before granting |
gitSnapshotIntervalMs | 30000 | checkpoint throttle |
auditLog | true | audit switch |
auditFile | ~/.dsh/permission-matrix/audit.jsonl | audit path |
Invalid configuration (unknown preset id, illegal tier) fails loudly at load time instead of silently degrading.