dsh-skillopt
Microsoft SkillOpt-Sleep 与 DeepSeek Harness 的集成：为你的 dsh 代理提供每晚睡眠周期，收集过往会话、重放重复任务，并在留出集门控后整合经过验证的技能。
插件会安装到这里；不确定时保持 web。
npx -y @deepseek-ai/dsh plugin --profile web add github:WODE25500/dsh-skillopt#130897a1a6dadf1e342ff63964abe60e940a01c6Skillopt 以 dsh-skillopt 发布，当前版本为 0.1.1。Plugin Hub 会校验它的 manifest，并保存精确安装来源，便于复现安装结果。
点击下方 Badge 复制 Markdown，粘贴到 README 即可。
验证 package.json 声明的 GitHub 仓库，即可管理这个公开页面。认领后，Hub 会优先安排当前版本的安全扫描，并在通过后公开展示结果。
认领这个 Plugin →Give your DeepSeek Harness agent a nightly sleep cycle: it reviews past
sessions offline, replays your recurring tasks on your own API budget, and
consolidates what it learns into validated skills behind a held-out gate. Same
engine as the Claude Code / Codex / Cursor integrations (skillopt_sleep),
wired into dsh's plugin system as native tools plus a bundled skill.
DeepSeek Harness is the "everything is a plugin" agent framework
(deepseek-ai/deepseek-harness).
Plugins are TypeScript modules exporting an apply(ctx) function that register
capabilities (tools, services, events, settings) on the Cordis context.
| Component | Purpose |
|---|---|
src/index.js | dsh plugin entry: registers 7 skillopt_* tools + Schemastery config |
cordis.patch.yml | bundle patch layer — drop dsh-skillopt into any profile's bundles |
skills/skillopt-sleep/SKILL.md | agent skill: when to use the tools, operating rules, data-boundary rules |
scripts/sleep.py | bootstrap/self-check runner (same command shape the tools use) |
package.json | npm package metadata (bundle manifest) |
| Tool | skillopt_sleep action | Behavior |
|---|---|---|
skillopt_status | status | state, engine availability, latest staged proposal & report |
skillopt_dry_run | dry-run | full preview (harvest+mine+replay), stages nothing |
skillopt_run | run | full cycle, stages a proposal (live files unchanged) |
skillopt_adopt | adopt | apply latest staged proposal (with backup) — the live-change boundary |
skillopt_harvest | harvest | read-only show/export of mined tasks |
skillopt_schedule / skillopt_unschedule | schedule / unschedule | install/remove the nightly cron entry |
pip install skillopt # or use this source checkout
Add dsh-skillopt to the profile's bundles, or in the profile cordis.patch.yml:
- insert:
- id: skillopt
name: './src/index.js'
config:
backend: mock # or codex / claude / cursor / pi / opencode / handoff …
project: /path/to/project
preferences: 'Always use async/await'
Run from a DeepSeek Harness source checkout (the official dev workflow,
pnpm resolves the workspace dsh bin):
pnpm dsh web --patch ./plugins/dsh/cordis.patch.yml
If dsh is installed globally (npm install -g), use it directly:
dsh web --patch ./plugins/dsh/cordis.patch.yml
Either way the patch inserts the skillopt plugin row into the profile; then
ask the agent: "Use skillopt_status to check the sleep cycle state."
| Key | Default | Purpose |
|---|---|---|
pythonCmd | python | Python interpreter for the engine |
module | — (bootstrap) | engine Python module override (python -m <module>) |
engineScript | — (scripts/sleep.py) | engine bootstrap script override |
project | — | default project directory |
scope | — | harvest scope: all | invoked |
backend | — | mock|claude|codex|copilot|cursor|pi|opencode|handoff|azure_openai |
source | — | claude|codex|copilot|cursor|pi|opencode|auto |
model | — | backend model override |
maxTasks / maxSessions | — | mine/harvest caps |
editBudget | — | bounded edits per cycle |
preferences | — | house rules for the reflection prior |
jsonOutput | false | machine-readable JSON output |
autoAdopt | false | OPERATOR-ONLY: auto-adopt a passed proposal without asking |
unscheduleAll | false | OPERATOR-ONLY: allow skillopt_unschedule to remove every managed entry |
timeoutMs | 600000 | per-call engine timeout in milliseconds |
Advanced engine keys (gate_mode, gate_metric, gate_no_regression,
dream_rollouts, recall_k, evolve_memory/evolve_skill) go in
~/.skillopt-sleep/config.json — the same file shared by all integrations.
mock/handoff make no network calls.run stages proposals; adopt is the normal live-change boundary and backs up first.skillopt_harvest
with output=), redact, set "reviewed": true, then replay — real backends
refuse unreviewed task files.python -m skillopt_sleep.experiments.run_experiment --persona researcher --assert-improves
See the SkillOpt-Sleep documentation for recorded results, limitations, and the supported integration surface.